Kyushu University Academic Staff Educational and Research Activities Database
List of Papers
Seiichi Uchida Last modified date:2020.06.29

Professor / Real World Robotics / Department of Advanced Information Technology / Faculty of Information Science and Electrical Engineering


Papers
1. Kana Aoki, Shinsuke Satoi, Shota Harada, Seiichi Uchida, Yoh Iwasa, Junichi Ikenouchi, Coordinated changes in cell membrane and cytoplasm during maturation of apoptotic bleb, Molecular biology of the cell, 10.1091/mbc.E19-12-0691, 31, 8, 833-844, 2020.04, Apoptotic cells form membrane blebs, but little is known about how the formation and dynamics of membrane blebs are regulated. The size of blebs gradually increases during the progression of apoptosis, eventually forming large extracellular vesicles called apoptotic bodies that have immune-modulating activities. In this study, we investigated the molecular mechanism involved in the differentiation of blebs into apoptotic blebs by comparing the dynamics of the bleb formed during cell migration and the bleb formed during apoptosis. We revealed that the enhanced activity of ROCK1 is required for the formation of small blebs in the early phase of apoptosis, which leads to the physical disruption of nuclear membrane and the degradation of Lamin A. In the late phase of apoptosis, the loss of asymmetry in phospholipids distribution caused the enlargement of blebs, which enabled translocation of damage-associated molecular patterns to the bleb cytoplasm and maturation of functional apoptotic blebs. Thus, changes in cell membrane dynamics are closely linked to cytoplasmic changes during apoptotic bleb formation..
2. Heon Song, Daiki Suehiro, Seiichi Uchida, Adaptive aggregation of arbitrary online trackers with a regret bound, 2020 IEEE/CVF Winter Conference on Applications of Computer Vision, WACV 2020 Proceedings - 2020 IEEE Winter Conference on Applications of Computer Vision, WACV 2020, 10.1109/WACV45572.2020.9093613, 670-678, 2020.03, We propose an online visual-object tracking method that is robust even in an adversarial environment, where various disturbances may occur on the target appearance, etc. The proposed method is based on a delayed-Hedge algorithm for aggregating multiple arbitrary online trackers with adaptive weights. The robustness in the tracking performance is guaranteed theoretically in term of "regret" by the property of the delayed-Hedge algorithm. Roughly speaking, the proposed method can achieve a similar tracking performance as the best one among all the trackers to be aggregated in an adversarial environment. The experimental study on various tracking tasks shows that the proposed method could achieve state-of-the-art performance by aggregating various online trackers..
3. Tomo Miyazaki, Tatsunori Tsuchiya, Yoshihiro Sugaya, Shinichiro Omachi, Masakazu Iwamura, Seiichi Uchida, Koichi Kise, Automatic Generation of Typographic Font from Small Font Subset, IEEE Computer Graphics and Applications, 10.1109/MCG.2019.2931431, 40, 1, 99-111, 2020.01, The automated generation of fonts containing a large number of characters is in high demand. For example, a typical Japanese font requires over 1000 characters. Unfortunately, professional typographers create the majority of fonts, resulting in significant financial and time investments for font generation. The main contribution of this article is the development of a method that automatically generates a target typographic font containing thousands of characters, from a small subset of character images in the target font. We generate characters other than the subset so that a complete font is obtained. We propose a novel font generation method with the capability to deal with various fonts, including a font composed of distinctive strokes, which are difficult for existing methods to handle. We demonstrated the proposed method by generating 2965 characters in 47 fonts. Moreover, objective and subjective evaluations verified that the generated characters are similar to the original characters..
4. Brian Kenji Iwana, Volkmar Frinken, Seiichi Uchida, DTW-NN
A novel neural network for time series recognition using dynamic alignment between inputs and weights, Knowledge-Based Systems, 10.1016/j.knosys.2019.104971, 188, 2020.01, This paper describes a novel model for time series recognition called a Dynamic Time Warping Neural Network (DTW-NN). DTW-NN is a feedforward neural network that exploits the elastic matching ability of DTW to dynamically align the inputs of a layer to the weights. This weight alignment replaces the standard dot product within a neuron with DTW. In this way, the DTW-NN is able to tackle difficulties with time series recognition such as temporal distortions and variable pattern length within a feedforward architecture. We demonstrate the effectiveness of DTW-NNs on four distinct datasets: online handwritten characters, accelerometer-based active daily life activities, spoken Arabic numeral Mel-Frequency Cepstrum Coefficients (MFCC), and one-dimensional centroid-radii sequences from leaf shapes. We show that the proposed method is an effective general approach to temporal pattern learning by achieving state-of-the-art results on these datasets..
5. Xiaotong Ji, Yuchen Zheng, Daiki Suehiro, Seiichi Uchida, Optimal Rejection Function Meets Character Recognition Tasks, 5th Asian Conference on Pattern Recognition, ACPR 2019 Pattern Recognition - 5th Asian Conference, ACPR 2019, Revised Selected Papers, 10.1007/978-3-030-41299-9_14, 169-183, 2020.01, In this paper, we propose an optimal rejection method for rejecting ambiguous samples by a rejection function. This rejection function is trained together with a classification function under the framework of Learning-with-Rejection (LwR). The highlights of LwR are: (1) the rejection strategy is not heuristic but has a strong background from a machine learning theory, and (2) the rejection function can be trained on an arbitrary feature space which is different from the feature space for classification. The latter suggests we can choose a feature space which is more suitable for rejection. Although the past research on LwR focused only its theoretical aspect, we propose to utilize LwR for practical pattern classification tasks. Moreover, we propose to use features from different CNN layers for classification and rejection. Our extensive experiments of notMNIST classification and character/non-character classification demonstrate that the proposed method achieves better performance than traditional rejection strategies..
6. Brian Kenji Iwana, Seiichi Uchida, Time series classification using local distance-based features in multi-modal fusion networks, Pattern Recognition, 10.1016/j.patcog.2019.107024, 97, 2020.01, We propose the use of a novel feature, called local distance features, for time series classification. The local distance features are extracted using Dynamic Time Warping (DTW) and classified using Convolutional Neural Networks (CNN). DTW is classically as a robust distance measure for distance-based time series recognition methods. However, by using DTW strictly as a global distance measure, information about the matching is discarded. We show that this information can further be used as supplementary input information in temporal CNNs. This is done by using both the raw data and the features extracted from DTW in multi-modal fusion CNNs. Furthermore, we explore the effects of different prototype selection methods, prototype numbers, and data fusion schemes induce on the accuracy. We perform experiments on a wide range of time series datasets including three Unipen handwriting datasets, four UCI Machine Learning Repository datasets, and 85 UCR Time Series Classification Archive datasets..
7. Hideaki Hayashi, Kohtaro Abe, Seiichi Uchida, GlyphGAN
Style-consistent font generation based on generative adversarial networks, Knowledge-Based Systems, 10.1016/j.knosys.2019.104927, 186, 2019.12, In this paper, we propose GlyphGAN: style-consistent font generation based on generative adversarial networks (GANs). GANs are a framework for learning a generative model using a system of two neural networks competing with each other. One network generates synthetic images from random input vectors, and the other discriminates between synthetic and real images. The motivation of this study is to create new fonts using the GAN framework while maintaining style consistency over all characters. In GlyphGAN, the input vector for the generator network consists of two vectors: character class vector and style vector. The former is a one-hot vector and is associated with the character class of each sample image during training. The latter is a uniform random vector without supervised information. In this way, GlyphGAN can generate an infinite variety of fonts with the character and style independently controlled. Experimental results showed that fonts generated by GlyphGAN have style consistency and diversity different from the training images without losing their legibility..
8. Brian Kenji Iwana, Ryohei Kuroki, Seiichi Uchida, Explaining convolutional neural networks using softmax gradient layer-wise relevance propagation, 17th IEEE/CVF International Conference on Computer Vision Workshop, ICCVW 2019 Proceedings - 2019 International Conference on Computer Vision Workshop, ICCVW 2019, 10.1109/ICCVW.2019.00513, 4176-4185, 2019.10, Convolutional Neural Networks (CNN) have become state-of-the-art in the field of image classification. However, not everything is understood about their inner representations. This paper tackles the interpretability and explainability of the predictions of CNNs for multi-class classification problems. Specifically, we propose a novel visualization method of pixel-wise input attribution called Softmax-Gradient Layer-wise Relevance Propagation (SGLRP). The proposed model is a class discriminate extension to Deep Taylor Decomposition (DTD) using the gradient of softmax to back propagate the relevance of the output probability to the input image. Through qualitative and quantitative analysis, we demonstrate that SGLRP can successfully localize and attribute the regions on input images which contribute to a target object's classification. We show that the proposed method excels at discriminating the target objects class from the other possible objects in the images. We confirm that SGLRP performs better than existing Layer-wise Relevance Propagation (LRP) based methods and can help in the understanding of the decision process of CNNs..
9. Yuchen Zheng, Brian Kenji Iwana, Seiichi Uchida, Mining the displacement of max-pooling for text recognition, Pattern Recognition, 10.1016/j.patcog.2019.05.014, 93, 558-569, 2019.09, The max-pooling operation in convolutional neural networks (CNNs)downsamples the feature maps of convolutional layers. However, in doing so, it loses some spatial information. In this paper, we extract a novel feature from pooling layers, called displacement features, and combine them with the features resulting from max-pooling to capture the structural deformations for text recognition tasks. The displacement features record the location of the maximal value in a max-pooling operation. Furthermore, we analyze and mine the class-wise trends of the displacement features. The extensive experimental results and discussions demonstrate that the proposed displacement features can improve the performance of the CNN based architectures and tackle the issues with the structural deformations of max-pooling in the text recognition tasks..
10. Yuchen Zheng, Wataru Ohyama, Brian Kenji Iwana, Seiichi Uchida, Capturing micro deformations from pooling layers for offline signature verification, 15th IAPR International Conference on Document Analysis and Recognition, ICDAR 2019 Proceedings - 15th IAPR International Conference on Document Analysis and Recognition, ICDAR 2019, 10.1109/ICDAR.2019.00180, 1111-1116, 2019.09, In this paper, we propose a novel Convolutional Neural Network (CNN) based method that extracts the location information (displacement features) of the maximums in the max-pooling operation and fuses it with the pooling features to capture the micro deformations between the genuine signatures and skilled forgeries as a feature extraction procedure. After the feature extraction procedure, we apply support vector machines (SVMs) as writer-dependent classifiers for each user to build the signature verification system. The extensive experimental results on GPDS-150, GPDS-300, GPDS-1000, GPDS-2000, and GPDS-5000 datasets demonstrate that the proposed method can discriminate the genuine signatures and their corresponding skilled forgeries well and achieve state-of-the-art results on these datasets..
11. Seokjun Kang, Brian Kenji Iwana, Seiichi Uchida, Cascading modular U-nets for document image binarization, 15th IAPR International Conference on Document Analysis and Recognition, ICDAR 2019 Proceedings - 15th IAPR International Conference on Document Analysis and Recognition, ICDAR 2019, 10.1109/ICDAR.2019.00113, 675-680, 2019.09, In recent years, U-Net has achieved good results in various image processing tasks. However, conventional U-Nets need to be re-trained for individual tasks with enough amount of images with ground-truth. This requirement makes U-Net not applicable to tasks with small amounts of data. In this paper, we propose to use 'modular' U-Nets, each of which is pre-trained to perform an existing image processing task, such as dilation, erosion, and histogram equalization. Then, to accomplish a specific image processing task, such as binarization of historical document images, the modular U-Nets are cascaded with inter-module skip connections and fine-tuned to the target task. We verified the proposed model using the Document Image Binarization Competition (DIBCO) 2017 dataset..
12. Xiaomeng Wu, Akisato Kimura, Brian Kenji Iwana, Seiichi Uchida, Kunio Kashino, Deep dynamic time warping
End-to-end local representation learning for online signature verification, 15th IAPR International Conference on Document Analysis and Recognition, ICDAR 2019 Proceedings - 15th IAPR International Conference on Document Analysis and Recognition, ICDAR 2019, 10.1109/ICDAR.2019.00179, 1103-1110, 2019.09, Siamese networks have been shown to be successful in learning deep representations for multivariate time series verification. However, most related studies optimize a global distance objective and suffer from a low discriminative power due to the loss of temporal information. To address this issue, we propose an end-to-end, neural network-based framework for learning local representations of time series, and demonstrate its effectiveness for online signature verification. This framework optimizes a Siamese network with a local embedding loss, and learns a feature space that preserves the temporal location-wise distances between time series. To achieve invariance to non-linear temporal distortion, we propose building a dynamic time warping block on top of the Siamese network, which will greatly improve the accuracy for local correspondences across intra-personal variability. Validation with respect to online signature verification demonstrates the advantage of our framework over existing techniques that use either handcrafted or learned feature representations..
13. Takuro Karamatsu, Daiki Suehiro, Seiichi Uchida, Logo design analysis by ranking, 15th IAPR International Conference on Document Analysis and Recognition, ICDAR 2019 Proceedings - 15th IAPR International Conference on Document Analysis and Recognition, ICDAR 2019, 10.1109/ICDAR.2019.00238, 1482-1487, 2019.09, In this paper, we analyze logo designs by using machine learning, as a promising trial of graphic design analysis. Specifically, we will focus on favicon images, which are tiny logos used as company icons on web browsers, and analyze them to understand their trends in individual industry classes. For example, if we can catch the subtle trends in favicons of financial companies, they will suggest to us how professional designers express the atmosphere of financial companies graphically. For the purpose, we will use top-rank learning, which is one of the recent machine learning methods for ranking and very suitable for revealing the subtle trends in graphic designs..
14. Taichi Sumi, Brian Kenji Iwana, Hideaki Hayashi, Seiichi Uchida, Modality conversion of handwritten patterns by cross variational autoencoders, 15th IAPR International Conference on Document Analysis and Recognition, ICDAR 2019 Proceedings - 15th IAPR International Conference on Document Analysis and Recognition, ICDAR 2019, 10.1109/ICDAR.2019.00072, 407-412, 2019.09, This research attempts to construct a network that can convert online and offline handwritten characters to each other. The proposed network consists of two Variational Auto-Encoders (VAEs) with a shared latent space. The VAEs are trained to generate online and offline handwritten Latin characters simultaneously. In this way, we create a cross-modal VAE (Cross-VAE). During training, the proposed Cross-VAE is trained to minimize the reconstruction loss of the two modalities, the distribution loss of the two VAEs, and a novel third loss called the space sharing loss. This third, space sharing loss is used to encourage the modalities to share the same latent space by calculating the distance between the latent variables. Through the proposed method mutual conversion of online and offline handwritten characters is possible. In this paper, we demonstrate the performance of the Cross-VAE through qualitative and quantitative analysis..
15. Kohei Baba, Seiichi Uchida, Brian Kenji Iwana, On the ability of a CNN to realize image-to-image language conversion, 15th IAPR International Conference on Document Analysis and Recognition, ICDAR 2019 Proceedings - 15th IAPR International Conference on Document Analysis and Recognition, ICDAR 2019, 10.1109/ICDAR.2019.00078, 448-453, 2019.09, The purpose of this paper is to reveal the ability that Convolutional Neural Networks (CNN) have on the novel task of image-to-image language conversion. We propose a new network to tackle this task by converting images of Korean Hangul characters directly into images of the phonetic Latin character equivalent. The conversion rules between Hangul and the phonetic symbols are not explicitly provided. The results of the proposed network show that it is possible to perform image-to-image language conversion. Moreover, it shows that it can grasp the structural features of Hangul even from limited learning data. In addition, it introduces a new network to use when the input and output have significantly different features..
16. Joonho Lee, Hideaki Hayashi, Wataru Ohyama, Seiichi Uchida, Page segmentation using a convolutional neural network with trainable co-occurrence features, 15th IAPR International Conference on Document Analysis and Recognition, ICDAR 2019 Proceedings - 15th IAPR International Conference on Document Analysis and Recognition, ICDAR 2019, 10.1109/ICDAR.2019.00167, 1023-1028, 2019.09, In document analysis, page segmentation is a fundamental task that divides a document image into semantic regions. In addition to local features, such as pixel-wise information, co-occurrence features are also useful for extracting texture-like periodic information for accurate segmentation. However, existing convolutional neural network (CNN)-based methods do not have any mechanisms that explicitly extract co-occurrence features. In this paper, we propose a method for page segmentation using a CNN with trainable multiplication layers (TMLs). The TML is specialized for extracting co-occurrences from feature maps, thereby supporting the detection of objects with similar textures and periodicities. This property is also considered to be effective for document image analysis because of regularity in text line structures, tables, etc. In the experiment, we achieved promising performance on a pixel-wise page segmentation task by combining TMLs with U-Net. The results demonstrate that TMLs can improve performance compared to the original U-Net. The results also demonstrate that TMLs are helpful for detecting regions with periodically repeating features, such as tables and main text..
17. Yan Zheng, Yuchen Zheng, Wataru Ohyama, Daiki Suehiro, Seiichi Uchida, RankSVM for offline signature verification, 15th IAPR International Conference on Document Analysis and Recognition, ICDAR 2019 Proceedings - 15th IAPR International Conference on Document Analysis and Recognition, ICDAR 2019, 10.1109/ICDAR.2019.00153, 928-933, 2019.09, Signature verification systems suffer from imbalanced learning, which imposes strict requirements on classifiers. The standard classification approaches, such as SVM, often degrade the performance for imbalanced data or require additional parameters for data balancing. In this study, as a new approach for signature verification, we use RankSVM as the writer-dependent classifiers, which theoretically guarantees the generalization performance for imbalanced data. To investigate the ability of RankSVM for solving imbalanced learning problems in signature verification tasks, the extensive experiments are conducted on bitmaps of GPDS-150, GPDS-300, GPDS-600, and GPDS-1000 datasets and deep features of GPDS-960 dataset. The experimental results demonstrate that the RankSVM-based approach obtains a nearly equivalent performance with the state-of-the-art method on deep features of the GPDS-960 dataset, and achieves significantly better performance than standard-SVM-based approach on bitmaps of GPDS-150, GPDS-300, GPDS-600, and GPDS-1000 datasets..
18. Toshiki Nakamura Nakamura, Anna Zhu, Seiichi Uchida, Scene text magnifier, 15th IAPR International Conference on Document Analysis and Recognition, ICDAR 2019 Proceedings - 15th IAPR International Conference on Document Analysis and Recognition, ICDAR 2019, 10.1109/ICDAR.2019.00137, 825-830, 2019.09, Scene text magnifier aims to magnify text in natural scene images without recognition. It could help the special groups, who have myopia or dyslexia to better understand the scene. In this paper, we design the scene text magnifier through interacted four CNN-based networks: character erasing, character extraction, character magnify, and image synthesis. The architecture of the networks are extended based on the hourglass encoderdecoders. It inputs the original scene text image and outputs the text magnified image while keeps the background unchange. Intermediately, we can get the side-output results of text erasing and text extraction. The four sub-networks are first trained independently and fine-tuned in end-to-end mode. The training samples for each stage are processed through a flow with original image and text annotation in ICDAR2013 and Flickr dataset as input, and corresponding text erased image, magnified text annotation, and text magnified scene image as output. To evaluate the performance of text magnifier, the Structural Similarity is used to measure the regional changes in each character region. The experimental results demonstrate our method can magnify scene text effectively without effecting the background..
19. Ryo Nakao, Brian Kenji Iwana, Seiichi Uchida, Selective super-resolution for scene text images, 15th IAPR International Conference on Document Analysis and Recognition, ICDAR 2019 Proceedings - 15th IAPR International Conference on Document Analysis and Recognition, ICDAR 2019, 10.1109/ICDAR.2019.00071, 401-406, 2019.09, In this paper, we realize the enhancement of super-resolution using images with scene text. Specifically, this paper proposes the use of Super-Resolution Convolutional Neural Networks (SRCNN) which are constructed to tackle issues associated with characters and text. We demonstrate that standard SRCNNs trained for general object super-resolution is not sufficient and that the proposed method is a viable method in creating a robust model for text. To do so, we analyze the characteristics of SRCNNs through quantitative and qualitative evaluations with scene text data. In addition, analysis using the correlation between layers by Singular Vector Canonical Correlation Analysis (SVCCA) and comparison of filters of each SRCNN using t-SNE is performed. Furthermore, in order to create a unified super-resolution model specialized for both text and objects, a model using SRCNNs trained with the different data types and Content-wise Network Fusion (CNF) is used. We integrate the SRCNN trained for character images and then SRCNN trained for general object images, and verify the accuracy improvement of scene images which include text. We also examine how each SRCNN affects super-resolution images after fusion..
20. Yuto Shinahara, Takuro Karamatsu, Daisuke Harada, Kota Yamaguchi, Seiichi Uchida, Serif or sans
Visual font analytics on book covers and online advertisements, 15th IAPR International Conference on Document Analysis and Recognition, ICDAR 2019 Proceedings - 15th IAPR International Conference on Document Analysis and Recognition, ICDAR 2019, 10.1109/ICDAR.2019.00170, 1041-1046, 2019.09, In this paper, we conduct a large-scale study of font statistics in book covers and online advertisements. Through the statistical study, we try to understand how graphic designers relate fonts and content genres and identify the relationship between font styles, colors, and genres. We propose an automatic approach to extract font information from graphic designs by applying a sequence of character detection, style classification, and clustering techniques to the graphic designs. The extracted font information is accumulated together with genre information, such as romance or business, for further trend analysis. Through our unique empirical study, we show that the collected font statistics reveal interesting trends in terms of how typographic design represents the impression and the atmosphere of the content genres..
21. Yosuke Onitsuka, Wataru Ohyama, Seiichi Uchida, Training convolutional autoencoders with metric learning, 15th IAPR International Conference on Document Analysis and Recognition, ICDAR 2019 Proceedings - 15th IAPR International Conference on Document Analysis and Recognition, ICDAR 2019, 10.1109/ICDAR.2019.00023, 86-91, 2019.09, We propose a new Training method that enables an autoencoder to extract more useful features for retrieval or classification tasks with limited-size datasets. Some targets in document analysis and recognition (DAR) including signature verification, historical document analysis, and scene text recognition, involve a common problem in which the size of the dataset available for training is small against the intra-class variety of the target appearance. Recently, several approaches, such as variational autoencoders and deep metric learning, have been proposed to obtain a feature representation that is suitable for the tasks. However, these methods sometimes cause an overfitting problem in which the accuracy of the test data is relatively low, while the performance for the training dataset is quite high. Our proposed method obtains feature representations for such tasks in DAR using convolutional autoencoders with metric learning. The accuracy is evaluated on an image-based retrieval of ancient Japanese signatures..
22. Shota Harada, Hideaki Hayashi, Ryoma Bise, Kiyohito Tanaka, Qier Meng, Seiichi Uchida, Endoscopic Image Clustering with Temporal Ordering Information Based on Dynamic Programming, 41st Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2019 2019 41st Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2019, 10.1109/EMBC.2019.8857011, 3681-3684, 2019.07, In this paper, we propose a clustering method with temporal ordering information for endoscopic image sequences. It is difficult to collect a sufficient amount of endoscopic image datasets to train machine learning techniques by manual labeling. The clustering of endoscopic images leads to group-based labeling, which is useful for reducing the cost of dataset construction. Therefore, in this paper, we propose a clustering method where the property of endoscopic image sequences is fully utilized. For the proposed method, a deep neural network was used to extract features from endoscopic images, and clustering with temporal ordering information was solved by dynamic programming. In the experiments, we clustered the esophagogastroduodenoscopy images. From the results, we confirmed that the performance was improved by using the sequential property..
23. Daisuke Harada, Ryoma Bise, Hiroki Tokunaga, Wataru Ohyama, Sanae Oka, Toshihiko Fujimori, Seiichi Uchida, Scribbles for Metric Learning in Histological Image Segmentation, 41st Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2019 2019 41st Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2019, 10.1109/EMBC.2019.8856465, 1026-1030, 2019.07, Segmentation is a fundamental process in biomedical image analysis that enables various types of analysis. Segmenting organs in histological microscopy images is problematic because the boundaries between regions are ambiguous, the images have various appearances, and the amount of training data is limited. To address these difficulties, supervised learning methods (e.g., convolutional neural networking (CNN)) are insufficient to predict regions accurately because they usually require a large amount of training data to learn the various appearances. In this paper, we propose a semi-automatic segmentation method that effectively uses scribble annotations for metric learning. Deep discriminative metric learning re-trains the representation of the feature space so that the distances between the samples with the same class labels are reduced, while those between ones with different class labels are enlarged. It makes pixel classification easy. Evaluation of the proposed method in a heart region segmentation task demonstrated that it performed better than three other methods..
24. Brian Kenji Iwana and Seiichi Uchida, Dynamic Weight Alignment for Temporal Convolutional Neural Networks, Proceedings of the 44th International Conference on Acoustics, Speech, and Signal Processing (ICASSP 2019, Brighton, UK), 2019.05.
25. Xiaomeng Wu, Akisato Kimura, Seiichi Uchida, and Kunio Kashino, Prewarping Siamese Network: Learning Local Representations For Online Signature Verification, Proceedings of the 44th International Conference on Acoustics, Speech, and Signal Processing (ICASSP 2019, Brighton, UK), 2019.05.
26. Brian Kenji Iwana, Seiichi Uchida, Dynamic Weight Alignment for Temporal Convolutional Neural Networks, 44th IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019 2019 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019 - Proceedings, 10.1109/ICASSP.2019.8682908, 3827-3831, 2019.05, In this paper, we propose a method of improving temporal Convolutional Neural Networks (CNN) by determining the optimal alignment of weights and inputs using dynamic programming. Conventional CNN convolutions linearly match the shared weights to a window of the input. However, it is possible that there exists a more optimal alignment of weights. Thus, we propose the use of Dynamic Time Warping (DTW) to dynamically align the weights to the input of the convolutional layer. Specifically, the dynamic alignment overcomes issues such as temporal distortion by finding the minimal distance matching of the weights and the inputs under constraints. We demonstrate the effectiveness of the proposed architecture on the Unipen online handwritten digit and character datasets, the UCI Spoken Arabic Digit dataset, and the UCI Activities of Daily Life dataset..
27. Xiaomeng Wu, Akisato Kimura, Seiichi Uchida, Kunio Kashino, Prewarping Siamese Network
Learning Local Representations for Online Signature Verification, 44th IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019 2019 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019 - Proceedings, 10.1109/ICASSP.2019.8683036, 2467-2471, 2019.05, We propose a neural network-based framework for learning local representations of multivariate time series, and demonstrate its effectiveness for online signature verification. In contrast to related works that optimize a global distance objective, we incorporate a Siamese network into dynamic time warping (DTW), leading to a novel prewarping Siamese network (PSN) optimized with a local embedding loss. PSN learns a feature space that preserves the temporal location-wise distances of local structures. Local embedding, along with the alignment conditions of DTW, imposes a temporal consistency constraint on the sequence-level distance measure while achieving invariance as regards non-linear distortions. Validation on online signature verification datasets demonstrates the advantage of our framework over existing techniques that use either handcrafted or learned feature representations..
28. Frédéric Rayar, Seiichi Uchida, Comic text detection using neural network approach, 25th International Conference on MultiMedia Modeling, MMM 2019 MultiMedia Modeling - 25th International Conference, MMM 2019, Proceedings, 10.1007/978-3-030-05716-9_60, 672-683, 2019.01, Text is a crucial element in comic books; hence text detection is a significant challenge in an endeavour to achieve comic processing. In this work, we study in what extent an off-the-shelf neural network approach for scene text detection can be used to perform comic text detection. Experiment on a public data set shows that such an approach allows to perform as well as methods of the literature, which is promising for building more accurate comic text detector in the future..
29. Hideaki Hayashi, Seiichi Uchida, A Trainable Multiplication Layer for Auto-correlation and Co-occurrence Extraction, 14th Asian Conference on Computer Vision, ACCV 2018 Computer Vision - ACCV 2018 - 14th Asian Conference on Computer Vision, Revised Selected Papers, 10.1007/978-3-030-20890-5_27, 414-430, 2019.01, In this paper, we propose a trainable multiplication layer (TML) for a neural network that can be used to calculate the multiplication between the input features. Taking an image as an input, the TML raises each pixel value to the power of a weight and then multiplies them, thereby extracting the higher-order local auto-correlation from the input image. The TML can also be used to extract co-occurrence from the feature map of a convolutional network. The training of the TML is formulated based on backpropagation with constraints to the weights, enabling us to learn discriminative multiplication patterns in an end-to-end manner. In the experiments, the characteristics of the TML are investigated by visualizing learned kernels and the corresponding output features. The applicability of the TML for classification and neural network interpretation is also evaluated using public datasets..
30. Shota Harada, Hideaki Hayashi, Seiichi Uchida, Biosignal Generation and Latent Variable Analysis with Recurrent Generative Adversarial Networks, IEEE Access, 10.1109/ACCESS.2019.2934928, 7, 144292-144302, 2019.01, The effectiveness of biosignal generation and data augmentation with biosignal generative models based on generative adversarial networks (GANs), which are a type of deep learning technique, was demonstrated in our previous paper. GAN-based generative models only learn the projection between a random distribution as input data and the distribution of training data. Therefore, the relationship between input and generated data is unclear, and the characteristics of the data generated from this model cannot be controlled. This study proposes a method for generating time-series data based on GANs and explores their ability to generate biosignals with certain classes and characteristics. Moreover, in the proposed method, latent variables are analyzed using canonical correlation analysis (CCA) to represent the relationship between input and generated data as canonical loadings. Using these loadings, we can control the characteristics of the data generated by the proposed method. The influence of class labels on generated data is analyzed by feeding the data interpolated between two class labels into the generator of the proposed GANs. The CCA of the latent variables is shown to be an effective method of controlling the generated data characteristics. We are able to model the distribution of the time-series data without requiring domain-dependent knowledge using the proposed method. Furthermore, it is possible to control the characteristics of these data by analyzing the model trained using the proposed method. To the best of our knowledge, this work is the first to generate biosignals using GANs while controlling the characteristics of the generated data..
31. Wataru Ohyama, Masakazu Suzuki, Seiichi Uchida, Detecting Mathematical Expressions in Scientific Document Images Using a U-Net Trained on a Diverse Dataset, IEEE Access, 10.1109/ACCESS.2019.2945825, 7, 144030-144042, 2019.01, A detection method for mathematical expressions in scientific document images is proposed. Inspired by the promising performance of U-Net, a convolutional network architecture originally proposed for the semantic segmentation of biomedical images, the proposed method uses image conversion by a U-Net framework. The proposed method does not use any information from mathematical and linguistic grammar so that it can be a supplemental bypass in the conventional mathematical optical character recognition (OCR) process pipeline. The evaluation experiments confirmed that (1) the performance of mathematical symbol and expression detection by the proposed method is superior to that of InftyReader, which is state-of-the-art software for mathematical OCR; (2) the coverage of the training dataset to the variation of document style is important; and (3) retraining with small additional training samples will be effective to improve the performance. An additional contribution is the release of a dataset for benchmarking the OCR for scientific documents..
32. Ryoma Bise, Kentaro Abe, Hideaki Hayashi, Kiyohito Tanaka, Seiichi Uchida, Efficient Soft-Constrained Clustering for Group-Based Labeling, 22nd International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2019 Medical Image Computing and Computer Assisted Intervention – MICCAI 2019 - 22nd International Conference, Proceedings, 10.1007/978-3-030-32254-0_47, 421-430, 2019.01, We propose a soft-constrained clustering method for group-based labeling of medical images. Since the idea of group-based labeling is to attach the label to a group of samples at once, we need to have groups (i.e., clusters) with high purity. The proposed method is formulated to achieve high purity even for difficult clustering tasks such as medical image clustering, where image samples of the same class are often very distant in their feature space. In fact, those images degrade the performance of conventional constrained clustering methods. Experiments with an endoscopy image dataset demonstrated that our method outperformed various state-of-the-art methods..
33. Daisuke Matsuoka, Masuo Nakano, Daisuke Sugiyama, Seiichi Uchida, Deep learning approach for detecting tropical cyclones and their precursors in the simulation by a cloud-resolving global nonhydrostatic atmospheric model, Progress in Earth and Planetary Science, 10.1186/s40645-018-0245-y, 5, 1, 2018.12, We propose a deep learning approach for identifying tropical cyclones (TCs) and their precursors. Twenty year simulated outgoing longwave radiation (OLR) calculated using a cloud-resolving global atmospheric simulation is used for training two-dimensional deep convolutional neural networks (CNNs). The CNNs are trained with 50,000 TCs and their precursors and 500,000 non-TC data for binary classification. Ensemble CNN classifiers are applied to 10 year independent global OLR data for detecting precursors and TCs. The performance of the CNNs is investigated for various basins, seasons, and lead times. The CNN model successfully detects TCs and their precursors in the western North Pacific in the period from July to November with a probability of detection (POD) of 79.9–89.1% and a false alarm ratio (FAR) of 32.8–53.4%. Detection results include 91.2%, 77.8%, and 74.8% of precursors 2, 5, and 7 days before their formation, respectively, in the western North Pacific. Furthermore, although the detection performance is correlated with the amount of training data and TC lifetimes, it is possible to achieve high detectability with a POD exceeding 70% and a FAR below 50% during TC season for several ocean basins, such as the North Atlantic, with a limited sample size and short lifetime. [Figure not available: see fulltext.]..
34. Hideaki Hayashi and Seiichi Uchida , A Trainable Multiplication Layer for Auto-correlation and Co-occurrence Extraction, Proceedings of the 14th Asian Conference on Computer Vision (ACCV 2018, Perth, Australia), 2018.12.
35. Yuchen Zheng, Brian Kenji Iwana, Seiichi Uchida, Discovering class-wise trends of max-pooling in subspace, 16th International Conference on Frontiers in Handwriting Recognition, ICFHR 2018 Proceedings - 2018 16th International Conference on Frontiers in Handwriting Recognition, ICFHR 2018, 10.1109/ICFHR-2018.2018.00026, 98-103, 2018.12, The traditional max-pooling operation in Convolutional Neural Networks (CNNs) only obtains the maximal value from a pooling window. However, it discards the information about the precise position of the maximal value. In this paper, we extract the location of the maximal value in a pooling window and transform it into 'displacement feature'. We analyze and discover the class-wise trend of the displacement features in many ways. The experimental results and discussion demonstrate that the displacement features have beneficial behaviors for solving the problems in max-pooling..
36. Kotaro Abe, Brian Kenji Iwana, Viktor Gosta Holmer, Seiichi Uchida, Font creation using class discriminative deep convolutional generative adversarial networks, 4th Asian Conference on Pattern Recognition, ACPR 2017 Proceedings - 4th Asian Conference on Pattern Recognition, ACPR 2017, 10.1109/ACPR.2017.99, 238-243, 2018.12, In this research, we attempt to generate fonts automatically using a modification of a Deep Convolutional Generative Adversarial Network (DCGAN) by introducing class consideration. DCGANs are the application of generative adversarial networks (GAN) which make use of convolutional and deconvolutional layers to generate data through adversarial detection. The conventional GAN is comprised of two neural networks that work in series. Specifically, it approaches an unsupervised method of data generation with the use of a generative network whose output is fed into a second discriminative network. While DCGANs have been successful on natural images, we show its limited ability on font generation due to the high variation of fonts combined with the need of rigid structures of characters. We propose a class discriminative DCGAN which uses a classification network to work alongside the discriminative network to refine the generative network. This results of our experiment shows a dramatic improvement over the conventional DCGAN..
37. Brian Kenji Iwana, Minoru Mori, Akisato Kimura, Seiichi Uchida, Introducing local distance-based features to temporal convolutional neural networks, 16th International Conference on Frontiers in Handwriting Recognition, ICFHR 2018 Proceedings - 2018 16th International Conference on Frontiers in Handwriting Recognition, ICFHR 2018, 10.1109/ICFHR-2018.2018.00025, 92-97, 2018.12, In this paper, we propose the use of local distance-based features determined by Dynamic Time Warping (DTW) for temporal Convolutional Neural Networks (CNN). Traditionally, DTW is used as a robust distance metric for time series patterns. However, this traditional use of DTW only utilizes the scalar distance metric and discards the local distances between the dynamically matched sequence elements. This paper proposes recovering these local distances, or DTW features, and utilizing them for the input of a CNN. We demonstrate that these features can provide additional information for the classification of isolated handwritten digits and characters. Furthermore, we demonstrate that the DTW features can be combined with the spatial coordinate features in multi-modal fusion networks to achieve state-of-the-art accuracy on the Unipen online handwritten character datasets..
38. Shailza Jolly, Brian Kenji Iwana, Ryohei Kuroki, Seiichi Uchida, How do Convolutional Neural Networks Learn Design?, 24th International Conference on Pattern Recognition, ICPR 2018 2018 24th International Conference on Pattern Recognition, ICPR 2018, 10.1109/ICPR.2018.8545624, 1085-1090, 2018.11, In this paper, we aim to understand the design principles in book cover images which are carefully crafted by experts. Book covers are designed in a unique way, specific to genres which convey important information to their readers. By using Convolutional Neural Networks (CNN) to predict book genres from cover images, visual cues which distinguish genres can be highlighted and analyzed. In order to understand these visual clues contributing towards the decision of a genre, we present the application of Layer-wise Relevance Propagation (LRP) on the book cover image classification results. We use LRP to explain the pixel-wise contributions of book cover design and highlight the design elements contributing towards particular genres. In addition, with the use of state-of-the-art object and text detection methods, insights about genre-specific book cover designs are discovered..
39. Shailza Jolly, Brian Kenji Iwana, Ryohei Kuroki, Seiichi Uchida, How do Convolutional Neural Networks Learn Design?, Proceedings of the 24th International Conference on Pattern Recognition (ICPR2018, Beijing, China), 2018.08.
40. Brian Kenji Iwana, Minoru Mori, Akisato Kimura and Seiichi Uchida, Introducing Local Distance-based Features to Temporal Convolutional Neural Networks , Proceedings of 16th International Conference on Frontiers in Handwriting Recognition (ICFHR2018, Niagara Falls, USA) , 2018.08.
41. Yuchen Zheng, Brian Kenji Iwana and Seiichi Uchida, Discovering Class-wise Trends of Max-pooling in Subspace, Proceedings of 16th International Conference on Frontiers in Handwriting Recognition (ICFHR2018, Niagara Falls, USA), 2018.08.
42. Maoko Tsukamoto, Kyoko Chiba, Yuriko Sobu, Yuzuha Shiraki, Yuka Okumura, Saori Hata, Akira Kitamura, Tadashi Nakaya, Seiichi Uchida, Masataka Kinjo, Hidenori Taru, Toshiharu Suzuki, The cytoplasmic region of the amyloid β-protein precursor (APP) is necessary and sufficient for the enhanced fast velocity of APP transport by kinesin-1, FEBS Letters, 10.1002/1873-3468.13204, 592, 16, 2716-2724, 2018.08, Amyloid β-protein precursor (APP) is transported mainly by kinesin-1 and at a higher velocity than other kinesin-1 cargos, such as Alcadein α (Alcα); this is denoted by the enhanced fast velocity (EFV). Interaction of the APP cytoplasmic region with kinesin-1, which is essential for EFV transport, is mediated by JNK-interacting protein 1 (JIP1). To determine the roles of interactions between the APP luminal region and cargo components, we monitored transport of chimeric cargo receptors, Alcα (luminal)–APP (cytoplasmic) and APP (luminal)–Alcα (cytoplasmic). Alcα-APP is transported at the EFV, whereas APP-Alcα is transported at the same velocity as wild-type Alcα. Thus, the cytoplasmic region of APP is necessary and sufficient for the EFV of APP transport by kinesin-1..
43. Frédéric Rayar, Seiichi Uchida, An Image-Based Representation for Graph Classification, Proceedings of IAPR Joint International Workshops on Statistical Techniques in Pattern Recognition and Structural and Syntactic Pattern Recognition (S+SSPR2018, Beijing, China), 2018.08.
44. Frédéric Rayar, Seiichi Uchida, On Fast Sample Preselection for Speeding up Convolutional Neural Network Training, Proceedings of IAPR Joint International Workshops on Statistical Techniques in Pattern Recognition and Structural and Syntactic Pattern Recognition (S+SSPR2018, Beijing, China), 2018.08.
45. Frederic Rayar, Masanori Goto, Seiichi Uchida, CNN training with graph-based sample preselection
application to handwritten character recognition, 13th IAPR International Workshop on Document Analysis Systems, DAS 2018 Proceedings - 13th IAPR International Workshop on Document Analysis Systems, DAS 2018, 10.1109/DAS.2018.10, 19-24, 2018.06, In this paper, we present a study on sample preselection in large training data set for CNN-based classification. To do so, we structure the input data set in a network representation, namely the Relative Neighbourhood Graph, and then extract some vectors of interest. The proposed preselection method is evaluated in the context of handwritten character recognition, by using two data sets, up to several hundred thousands of images. It is shown that the graph-based preselection can reduce the training data set without degrading the recognition accuracy of a non pretrained CNN shallow model..
46. Gantugs Atarsaikhan, Brian Kenji Iwana, Seiichi Uchida, Contained neural style transfer for decorated logo generation, 13th IAPR International Workshop on Document Analysis Systems, DAS 2018 Proceedings - 13th IAPR International Workshop on Document Analysis Systems, DAS 2018, 10.1109/DAS.2018.78, 317-322, 2018.06, Making decorated logos requires image editing skills, without sufficient skills, it could be a time-consuming task. While there are many on-line web services to make new logos, they have limited designs and duplicates can be made. We propose using neural style transfer with clip art and text for the creation of new and genuine logos. We introduce a new loss function based on distance transform of the input image, which allows the preservation of the silhouettes of text and objects. The proposed method contains style transfer to only a designated area. We demonstrate the characteristics of proposed method. Finally, we show the results of logo generation with various input images..
47. Liuan Wang, Jun Sun, Seiichi Uchida, Text line extraction based on integrated k-shortest paths optimization, 13th IAPR International Workshop on Document Analysis Systems, DAS 2018 Proceedings - 13th IAPR International Workshop on Document Analysis Systems, DAS 2018, 10.1109/DAS.2018.68, 85-90, 2018.06, Text in images can be utilized in many image understanding applications due to the exact semantic information. In this paper, we propose a novel integrated k-shortest paths optimization based text line extraction method. Firstly, the candidate text components are extracted by the Maximal Stable Extremal Region (MSER) algorithm on gray, red, green and blue channels. Secondly, one integrated directed graph on red, green, and blue channels are constructed upon the candidate text components, which can effectively incorporate different channels into one framework. Then, the integrated directed graph is transformed guided by the extracted text lines in gray channel to reduced the computational complexity. Finally, we use the k-shortest paths optimization algorithm to extract the text lines by taking advantage of the particular structure of the integrated directed graph. Experimental results demonstrate the effectiveness of the proposed method in comparison with state-of-the-art methods..
48. Anna Zhu, Seiichi Uchida, Scene word recognition from pieces to whole, Frontiers of Computer Science, 10.1007/s11704-017-6420-2, 1-10, 2018.04, Convolutional neural networks (CNNs) have had great success with regard to the object classification problem. For character classification, we found that training and testing using accurately segmented character regions with CNNs resulted in higher accuracy than when roughly segmented regions were used. Therefore, we expect to extract complete character regions from scene images. Text in natural scene images has an obvious contrast with its attachments. Many methods attempt to extract characters through different segmentation techniques. However, for blurred, occluded, and complex background cases, those methods may result in adjoined or over segmented characters. In this paper, we propose a scene word recognition model that integrates words from small pieces to entire after-cluster-based segmentation. The segmented connected components are classified as four types: background, individual character proposals, adjoined characters, and stroke proposals. Individual character proposals are directly inputted to a CNN that is trained using accurately segmented character images. The sliding window strategy is applied to adjoined character regions. Stroke proposals are considered as fragments of entire characters whose locations are estimated by a stroke spatial distribution system. Then, the estimated characters from adjoined characters and stroke proposals are classified by a CNN that is trained on roughly segmented character images. Finally, a lexicondriven integration method is performed to obtain the final word recognition results. Compared to other word recognition methods, our method achieves a comparable performance on Street View Text and the ICDAR 2003 and ICDAR Received August 22, 2016; accepted March 12, 2017 E-mail: annakkk@live.com 2013 benchmark databases. Moreover, our method can deal with recognizing text images of occlusion and improperly segmented text images..
49. Frédéric Rayar, Masanori Goto and Seiichi Uchida, CNN Training with Graph-Based Sample Preselection: Application to Handwritten Character Recognition, Proceedings of The 13th IAPR International Workshop on Document Analysis Systems (DAS2018, Viena, Austria), 2018.04.
50. Liuan Wang, Jun Sun and Seiichi Uchida, Text Line Extraction based on Integrated K-shortest Paths Optimization, Proceedings of The 13th IAPR International Workshop on Document Analysis Systems (DAS2018, Viena, Austria), 2018.04.
51. Gantugs Atarsaikhan, Brian Kenji Iwana and Seiichi Uchida, Contained Neural Style Transfer for Decorated Logo Generation, Proceedings of The 13th IAPR International Workshop on Document Analysis Systems (DAS2018, Viena, Austria), 2018.04.
52. Liuan Wang, Seiichi Uchida, Anna Zhu, Jun Sun, Human Reading Knowledge Inspired Text Line Extraction, Cognitive Computation, 10.1007/s12559-017-9490-4, 10, 1, 84-93, 2018.02, Text in images contains exact semantic information and the text knowledge can be utilized in many image cognition and understanding applications. The human reading habits can provide the clues of text line structure for text line extraction. In this paper, we propose a novel human reading knowledge inspired text line extraction method based on k-shortest paths global optimization. Firstly, the candidate character extraction is reformulated as Maximal Stable Extremal Region (MSER) algorithm on gray, red, blue, and green channels of the target images, and the extracted MSERs are fed into Convolutional Neural Network (CNN) to remove the noise components. Then, the directed graph is built upon the character component nodes with edges inspired by human reading sense. The directed graph can automatically construct the relationship to eliminate the disorder of candidate text components. The text line paths optimization is inspired by the human reading ability in planning of a text line path sequentially. Therefore, the text line extraction problem can be solved using the k-shortest paths optimization algorithm by taking advantage of the human reading sense structure of the directed graph. It can extract the text lines iteratively to avoid the exhaustive searching and obtain global optimized text line number. The proposed method achieves the f-measure of 0.820 and 0.812 on public ICDAR2011 and ICDAR2013 dataset, respectively. The experimental results demonstrate the effectiveness of the proposed human reading knowledge inspired text line extraction method in comparison with state-of-the-art methods This paper presents one human reading knowledge inspired text line extraction method, which approves that the human reading knowledge can benefit the text line extraction and image text discovery..
53. Brian Kenji Iwana, Letao Zhou, Kumiko Tanaka-Ishii, Seiichi Uchida, Component Awareness in Convolutional Neural Networks, 14th IAPR International Conference on Document Analysis and Recognition, ICDAR 2017 Proceedings - 14th IAPR International Conference on Document Analysis and Recognition, ICDAR 2017, 10.1109/ICDAR.2017.72, 1, 394-399, 2018.01, In this work, we investigate the ability of Convolutional Neural Networks (CNN) to infer the presence of components that comprise an image. In recent years, CNNs have achieved powerful results in classification, detection, and segmentation. However, these models learn from instance-level supervision of the detected object. In this paper, we determine if CNNs can detect objects using image-level weakly supervised labels without localization. To demonstrate that a CNN can infer awareness of objects, we evaluate a CNN's classification ability with a database constructed of Chinese characters with only character-level labeled components. We show that the CNN is able to achieve a high accuracy in identifying the presence of these components without specific knowledge of the component. Furthermore, we verify that the CNN is deducing the knowledge of the target component by comparing the results to an experiment with the component removed. This research is important for applications with large amounts of data without robust annotation such as Chinese character recognition..
54. Jinho Lee, Brian Kenji Iwana, Shouta Ide, Hideaki Hayashi, Seiichi Uchida, Globally Optimal Object Tracking with Complementary Use of Single Shot Multibox Detector and Fully Convolutional Network, 8th Pacific-Rim Symposium on Image and Video Technology, PSIVT 2017 Image and Video Technology - 8th Pacific-Rim Symposium, PSIVT 2017, Revised Selected Papers, 10.1007/978-3-319-75786-5_10, 110-122, 2018.01, Object tracking is one of the most important but still difficult tasks in computer vision and pattern recognition. The main difficulties in the tracking task are appearance variation of target objects and occlusion. To deal with those difficulties, we propose a object tracking method combining Single Shot Multibox Detector (SSD), Fully Convolutional Network (FCN) and Dynamic Programming (DP). SSD and FCN provide a probability value of the target object which allows for appearance variation within each category. DP provides a globally optimal tracking path even with severe occlusions. Through several experiments, we confirmed that their combination realized a robust object tracking method. Also, in contrast to traditional trackers, initial position and a template of the target do not need to be specified. We show that the proposed method has a higher performance than the traditional trackers in tracking various single objects through video frames..
55. Shota Ide, Seiichi Uchida, How Does a CNN Manage Different Printing Types?, 14th IAPR International Conference on Document Analysis and Recognition, ICDAR 2017 Proceedings - 14th IAPR International Conference on Document Analysis and Recognition, ICDAR 2017, 10.1109/ICDAR.2017.167, 1, 1004-1009, 2018.01, In past OCR research, different OCR engines are used for different printing types, i.e., machine-printed characters, handwritten characters, and decorated fonts. A recent research, however, reveals that convolutional neural networks (CNN) can realize a universal OCR, which can deal with any printing types without pre-classification into individual types. In this paper, we analyze how CNN for universal OCR manage the different printing types. More specifically, we try to find where a handwritten character of a class and a machine-printed character of the same class are 'fused' in CNN. For analysis, we use two different approaches. The first approach is statistical analysis for detecting the CNN units which are sensitive (or insensitive) to type difference. The second approach is network-based visualization of pattern distribution in each layer. Both analyses suggest the same trend that types are not fully fused in convolutional layers but the distributions of the same class from different types become closer in upper layers..
56. Gantugs Atarsaikhan, Brian Kenji Iwana, Atsushi Narusawa, Keiji Yanai, Seiichi Uchida, Neural Font Style Transfer, 1st Workshop of Machine Learning under International Conference on Document Analysis and Recognition, ICDAR-WML 2017 Proceedings - 1st Workshop of Machine Learning under International Conference on Document Analysis and Recognition, ICDAR-WML 2017, 10.1109/ICDAR.2017.328, 5, 51-56, 2018.01, In this paper, we chose an approach to generate fonts by using neural style transfer. Neural style transfer uses Convolution Neural Networks(CNN) to transfer the style of one image to another. By modifying neural style transfer, we can achieve neural font style transfer. We also demonstrate the effects of using different weighted factors, character placements, and orientations. In addition, we show the results of using non-Latin alphabets, non-text patterns, and non-text images as style images. Finally, we provide insight into the characteristics of style transfer with fonts..
57. Toshiki Nakamura, Anna Zhu, Keiji Yanai, Seiichi Uchida, Scene Text Eraser, 14th IAPR International Conference on Document Analysis and Recognition, ICDAR 2017 Proceedings - 14th IAPR International Conference on Document Analysis and Recognition, ICDAR 2017, 10.1109/ICDAR.2017.141, 1, 832-837, 2018.01, The character information in natural scene images contains various personal information, such as telephone numbers, home addresses, etc. It is a high risk of leakage the information if they are published. In this paper, we proposed a scene text erasing method to properly hide the information via an inpainting convolutional neural network (CNN) model. The input is a scene text image, and the output is expected to be text erased image with all the character regions filled up the colors of the surrounding background pixels. This work is accomplished byaCNNmodelthroughconvolutiontodeconvolutionwithinterconnection process. The training samples and the corresponding inpainting images are considered as teaching signals for training. To evaluate the text erasing performance, the output images are detected by a novel scene text detection method. Subsequently, the same measurement on text detection is utilized for testing the images in benchmark dataset ICDAR2013. Compared with direct text detection way, the scene text erasing process demonstrates a drastically decrease on the precision, recall and f-score. That proves the effectiveness of proposed method for erasing the text in natural scene images..
58. Anna Zhu, Seiichi Uchida, Scene Text Relocation with Guidance, 14th IAPR International Conference on Document Analysis and Recognition, ICDAR 2017 Proceedings - 14th IAPR International Conference on Document Analysis and Recognition, ICDAR 2017, 10.1109/ICDAR.2017.212, 1, 1289-1294, 2018.01, Applying object proposal technique for scene text detection becomes popular for its significant improvement in speed and accuracy for object detection. However, some of the text regions after the proposal classification are overlapped and hard to remove or merge. In this paper, we present a scene text relocation system that refines the detection from text proposals to text. An object proposal-based deep neural network is employed to get the text proposals. To tackle the detection overlapping problem, a refinement deep neural network relocates the overlapped regions by estimating the text probability inside, and locating the accurate text regions by thresholding. Since the spacebetweenwordsindifferenttextlinesarevarious, aguidance mechanism is proposed in text relocation to guide where to extract the text regions in word level. This refinement procedure helps boost the precision after removing multiple overlapped text regions or joint cracked text regions. The experimental results on standard benchmark ICDAR 2013 demonstrate the effectiveness of the proposed approach..
59. Koichi Kise, Shinichiro Omachi, Seiichi Uchida, Masakazu Iwamura, Welcome Message from the ICDAR 2017 General and Executive Chairs, 14th IAPR International Conference on Document Analysis and Recognition, ICDAR 2017 Proceedings of the International Conference on Document Analysis and Recognition, ICDAR, 10.1109/ICDAR.2017.5, 1, xxiv-xxv, 2018.01.
60. Frédéric Rayar, Seiichi Uchida, An image-based representation for graph classification, Joint IAPR International Workshops on Structural and Syntactic Pattern Recognition, SSPR 2018 and Statistical Techniques in Pattern Recognition, SPR 2018 Structural, Syntactic, and Statistical Pattern Recognition - Joint IAPR International Workshop, S+SSPR 2018, Proceedings, 10.1007/978-3-319-97785-0_14, 140-149, 2018.01, This paper proposes to study the relevance of image representations to perform graph classification. To do so, the adjacency matrix of a given graph is reordered using several matrix reordering algorithms. The resulting matrix is then converted into an image thumbnail, that is used to represent the graph. Experimentation on several chemical graph data sets and an image data set show that the proposed graph representation performs as well as the state-of-the-art methods..
61. Frédéric Rayar, Seiichi Uchida, On fast sample preselection for speeding up convolutional neural network training, Joint IAPR International Workshops on Structural and Syntactic Pattern Recognition, SSPR 2018 and Statistical Techniques in Pattern Recognition, SPR 2018 Structural, Syntactic, and Statistical Pattern Recognition - Joint IAPR International Workshop, S+SSPR 2018, Proceedings, 10.1007/978-3-319-97785-0_7, 65-75, 2018.01, We propose a fast hybrid statistical and graph-based sample preselection method for speeding up CNN training process. To do so, we process each class separately: some candidates are first extracted based on their distances to the class mean. Then, we structure all the candidates in a graph representation and use it to extract the final set of preselected samples. The proposed method is evaluated and discussed based on an image classification task, on three data sets that contain up to several hundred thousands of images..
62. Kotaro Abe, Brian Kenji Iwana, Viktor Gösta Holmér and Seiichi Uchida, Font Creation Using Generative Adversarial Networks with Class Discrimination, Proceedings of Asian Conference on Pattern Recognition (ACPR2017, Nanjing, China), 2017.10.
63. Tomohiro Nakayasu, Masaki Yasugi, Soma Shiraishi, Seiichi Uchida, Eiji Watanabe, Three-Dimensional Computer Graphic Animations for Studying Social Approach Behaviour in Medaka Fish: Effects of Systematic Manipulation of Morphological and Motion Cues , PLoS ONE, 2017.04.
64. Brian K. Iwana, Kaspar Riesen, Volkmar Frinken, Seiichi Uchida, Efficient Temporal Pattern Recognition by Means of Dissimilarity Space Embedding with Discriminative Prototypes , Pattern Recognition, 2017.04.
65. Koichi Kise, Shinichiro Omachi, Seiichi Uchida, Masakazu Iwamura, Masahiko Inami, Kai Kunze, Reading-life log as a new paradigm of utilizing character and document media, Horizontal Expansion, 10.1007/978-4-431-56535-2_7, 2, 197-233, 2017.04, "You are what you read." As this sentence implies, reading is important for building our minds. We are investing a huge amount of time for reading to input information. However the activity of "reading" is done only by each individual in an analog way and nothing is digitally recorded and reused. In order to solve this problem, we record reading activities as digital data and analyze them for various goals. We call this research "reading-life log." In this chapter, we describe our achievements of the reading-life log. A target of the reading-life log is to analyze reading activities quantitatively and qualitatively: when, how much, what you read, and how you read in terms of your interests and understanding. Body-worn sensors including intelligent eyewear are employed for this purpose. Another target is to analyze the contents of documents based on the users' reading activities: for example, which are the parts most people feel difficult/interesting. Materials to be read are not limited to books and documents. Scene texts are also important materials which guide human activities..
66. Brian Kenji Iwana, Volkmar Frinken, Kaspar Riesen, Seiichi Uchida, Efficient temporal pattern recognition by means of dissimilarity space embedding with discriminative prototypes, Pattern Recognition, 10.1016/j.patcog.2016.11.013, 64, 268-276, 2017.04, Dissimilarity space embedding (DSE) presents a method of representing data as vectors of dissimilarities. This representation is interesting for its ability to use a dissimilarity measure to embed various patterns (e.g. graph patterns with different topology and temporal patterns with different lengths) into a vector space. The method proposed in this paper uses a dynamic time warping (DTW) based DSE for the purpose of the classification of massive sets of temporal patterns. However, using large data sets introduces the problem of requiring a high computational cost. To address this, we consider a prototype selection approach. A vector space created by DSE offers us the ability to treat its independent dimensions as features allowing for the use of feature selection. The proposed method exploits this and reduces the number of prototypes required for accurate classification. To validate the proposed method we use two-class classification on a data set of handwritten on-line numerical digits. We show that by using DSE with ensemble classification, high accuracy classification is possible with very few prototypes..
67. Tomohiro Nakayasu, Masaki Yasugi, Soma Shiraishi, Seiichi Uchida, Eiji Watanabe, Three-dimensional computer graphic animations for studying social approach behaviour in medaka fish
Effects of systematic manipulation of morphological and motion cues, PloS one, 10.1371/journal.pone.0175059, 12, 4, 2017.04, We studied social approach behaviour in medaka fish using three-dimensional computer graphic (3DCG) animations based on the morphological features and motion characteristics obtained from real fish. This is the first study which used 3DCG animations and examined the relative effects of morphological and motion cues on social approach behaviour in medaka. Various visual stimuli, e.g., lack of motion, lack of colour, alternation in shape, lack of locomotion, lack of body motion, and normal virtual fish in which all four features (colour, shape, locomotion, and body motion) were reconstructed, were created and presented to fish using a computer display. Medaka fish presented with normal virtual fish spent a long time in proximity to the display, whereas time spent near the display was decreased in other groups when compared with normal virtual medaka group. The results suggested that the naturalness of visual cues contributes to the induction of social approach behaviour. Differential effects between body motion and locomotion were also detected. 3DCG animations can be a useful tool to study the mechanisms of visual processing and social behaviour in medaka..
68. Kenji Kimura, Alexandre Mamane, Tohru Sasaki, Kohta Sato, Jun Takagi, Ritsuya Niwayama, Lars Hufnagel, Yuta Shimamoto, Jean-François Joanny, Seiichi Uchida, Akatsuki Kimura, Endoplasmic-Reticulum-Mediated Microtubule Alignment Governs Cytoplasmic Streaming , Nature Cell Biology, 2017.03.
69. Yuki Sato, Kei Nagatoshi, Ayumi Hamano, Yuko Imamura, David Huss, Seiichi Uchida, Rusty Lansford, Basal Filopodia and Vascular Mechanical Stress Organize Fibronectin into Pillars Bridging the Mesoderm-Endoderm Gap , Development, 2017.02.
70. Masanori Goto, Ryosuke Ishida, Seiichi Uchida, A Preselection-Based Fast Support Vector Machine Learning for Large-Scale Pattern Sets using Compressed Relative Neighborhood Graph, Research Reports on Information Science and Electrical Engineering of Kyushu University, 22, 1, 1-7, 2017.01, We propose a pre-selection method for training support vector machines (SVM) with a largescale dataset. Specifically, the proposed method selects patterns around the class boundary and the selected data is fed to train an SVM. For the selection, that is, searching for boundary patterns, we utilize a compressed representation of relative neighborhood graph (Clustered-RNG). A Clustered-RNG is a network of neighboring patterns which have a different class label and thus, we can find boundary patterns between different classes. Through large-scale handwritten digit pattern recognition experiments, we show that the proposed pre-selection method accelerates SVM training process 10 times faster without degrading recognition accuracy..
71. Seiichi Uchida, Shota Ide, Brian Kenji Iwana, Anna Zhu, A further step to perfect accuracy by training CNN with larger data, 15th International Conference on Frontiers in Handwriting Recognition, ICFHR 2016 Proceedings - 2016 15th International Conference on Frontiers in Handwriting Recognition, ICFHR 2016, 10.1109/ICFHR.2016.0082, 405-410, 2017.01, Convolutional Neural Networks (CNN) are on the forefront of accurate character recognition. This paper explores CNNs at their maximum capacity by implementing the use of large datasets. We show a near-perfect performance by using a dataset of about 820,000 real samples of isolated handwritten digits, much larger than the conventional MNIST database. In addition, we report a near-perfect performance on the recognition of machine-printed digits and multi-font digital born digits. Also, in order to progress toward a universal OCR, we propose methods of combining the datasets into one classifier. This paper reveals the effects of combining the datasets prior to training and the effects of transfer learning during training. The results of the proposed methods also show an almost perfect accuracy suggesting the ability of the network to generalize all forms of text..
72. Brian Kenji Iwana, Volkmar Frinken, Seiichi Uchida, A robust dissimilarity-based neural network for temporal pattern recognition, 15th International Conference on Frontiers in Handwriting Recognition, ICFHR 2016 Proceedings - 2016 15th International Conference on Frontiers in Handwriting Recognition, ICFHR 2016, 10.1109/ICFHR.2016.0058, 265-270, 2017.01, Temporal pattern recognition is challenging because temporal patterns require extra considerations over other data types, such as order, structure, and temporal distortions. Recently, there has been a trend in using large data and deep learning, however, many of the tools cannot be directly used with temporal patterns. Convolutional Neural Networks (CNN) for instance are traditionally used for visual and image pattern recognition. This paper proposes a method using a neural network to classify isolated temporal patterns directly. The proposed method uses dynamic time warping (DTW) as a kernel-like function to learn dissimilarity-based feature maps as the basis of the network. We show that using the proposed DTW-NN, efficient classification of on-line handwritten digits is possible with accuracies comparable to state-of-the-art methods..
73. Yuki Sato, Kei Nagatoshi, Ayumi Hamano, Yuko Imamura, David Huss, Seiichi Uchida, Rusty Lansford, Basal filopodia and vascular mechanical stress organize fibronectin into pillars bridging the mesoderm-endoderm gap, Development (Cambridge, England), 10.1242/dev.141259, 144, 2, 281-291, 2017.01, Cells may exchange information with other cells and tissues by exerting forces on the extracellular matrix (ECM). Fibronectin (FN) is an important ECM component that forms fibrils through cell contacts and creates directionally biased geometry. Here, we demonstrate that FN is deposited as pillars between widely separated germ layers, namely the somitic mesoderm and the endoderm, in quail embryos. Alongside the FN pillars, long filopodia protrude from the basal surfaces of somite epithelial cells. Loss-of-function of Ena/VASP, α5β1-integrins or talin in the somitic cells abolished the FN pillars, indicating that FN pillar formation is dependent on the basal filopodia through these molecules. The basal filopodia and FN pillars are also necessary for proper somite morphogenesis. We identified a new mechanism contributing to FN pillar formation by focusing on cyclic expansion of adjacent dorsal aorta. Maintenance of the directional alignment of the FN pillars depends on pulsatile blood flow through the dorsal aortae. These results suggest that the FN pillars are specifically established through filopodia-mediated and pulsating force-related mechanisms..
74. Shigeru Takano, Maiya Hori, Takayuki Goto, Seiichi Uchida, Ryo Kurazume, Rin Ichiro Taniguchi, Deep learning-based prediction method for people flows and their anomalies, 6th International Conference on Pattern Recognition Applications and Methods, ICPRAM 2017 ICPRAM 2017 - Proceedings of the 6th International Conference on Pattern Recognition Applications and Methods, 2017-January, 676-683, 2017.01, This paper proposes prediction methods for people flows and anomalies in people flows on a university campus. The proposed methods are based on deep learning frameworks. By predicting the statistics of people flow conditions on a university campus, it becomes possible to create applications that predict future crowded places and the time when congestion will disappear. Our prediction methods will be useful for developing applications for solving problems in cities..
75. Kenji Kimura, Alexandre Mamane, Tohru Sasaki, Kohta Sato, Jun Takagi, Ritsuya Niwayama, Lars Hufnagel, Yuta Shimamoto, Jean François Joanny, Seiichi Uchida, Akatsuki Kimura, Endoplasmic-reticulum-mediated microtubule alignment governs cytoplasmic streaming, Nature Cell Biology, 10.1038/ncb3490, 19, 4, 399-406, 2017.01, Cytoplasmic streaming refers to a collective movement of cytoplasm observed in many cell types1-7. The mechanism of meiotic cytoplasmic streaming (MeiCS) in Caenorhabditis elegans zygotes is puzzling as the direction of the flow is not predefined by cell polarity and occasionally reverses6. Here, we demonstrate that the endoplasmic reticulum (ER) network structure is required for the collective flow. Using a combination of RNAi, microscopy and image processing of C. elegans zygotes, we devise a theoretical model, which reproduces and predicts the emergence and reversal of the flow. We propose a positive-feedback mechanism, where a local flow generated along a microtubule is transmitted to neighbouring regions through the ER. This, in turn, aligns microtubules over a broader area to self-organize the collective flow. The proposed model could be applicable to various cytoplasmic streaming phenomena in the absence of predefined polarity. The increased mobility of cortical granules by MeiCS correlates with the efficient exocytosis of the granules to protect the zygotes from osmotic and mechanical stresses..
76. Masanori Goto, Ryosuke Ishida, Seiichi Uchida, Visualizing the distribution of a large-scale pattern set using compressed relative neighborhood graph, IEEJ Transactions on Electronics, Information and Systems, 10.1541/ieejeiss.137.1495, 137, 11, 1495-1505, 2017.01, The goal of this research is to understand the true distribution of character patterns. Advances in computer technology for mass storage and digital processing have paved way to process a massive dataset for various pattern recognition problems. If we can represent and analyze the distribution of a large-scale pattern set directly and understand its relationships deeply, it should be helpful for improving classifier for pattern recognition. For this purpose, we use a visualization method to represent the distribution of patterns using a relative neighborhood graph (RNG), where each node corresponds to a single pattern. Specifically, we visualize the pattern distribution using a compressed representation of RNG (Clustered-RNG). Clustered-RNG can visualize inter-class relationships (e.g. neighboring relationships and overlaps of pattern distribution among "multiple classes") and it represents the distribution of the patterns without any assumption, approximation or loss. Through large-scale printed and handwritten digit pattern experiments, we show the properties and validity of the visualization using Clustered-RNG..
77. Seiichi Uchida and Yuto Shinahara, What Does Scene Text Tell Us?, Proceedings of the 23rd International Conference on Pattern Recognition, 2016.12.
78. Brian Iwana, Seiichi Uchida and Volkmar Frinken, A Robust Dissimilarity-based Neural Network for Temporal Pattern Recognition, Proceedings of the 15th International Conference on Frontiers in Handwriting Recognition, 2016.10.
79. Seiichi Uchida, Shota Ide, Brian Iwana and Anna Zhu, A Further Step to Perfect Accuracy by Training CNN with Larger Data, Proceedings of the 15th International Conference on Frontiers in Handwriting Recognition, 2016.10.
80. Zhu Anna, Renwu Gao, Seiichi Uchida, Could Scene Context be Beneficial for Scene Text Detection?, Pattern Recognition, 10.1016/j.patcog.2016.04.011, 2016.10.
81. O. Nedzvedz, S. Ablameyko, Seiichi Uchida, Extraction and tracking living cells in medical images, 2016 International Conference on Information and Digital Technologies, IDT 2016 IDT 2016 - Proceedings of the International Conference on Information and Digital Technologies 2016, 10.1109/DT.2016.7557173, 198-202, 2016.08, One of the important problems of cytological image analysis is the cell segmentation. Today the most perspective direction of cytological image analysis is living cells investigation. Such images lead to many troubles for cell analysis. In this paper, we propose a solution of one such problems: pattern extraction of living cells from its aggregation and measurement of their 3D characteristics..
82. Shigeru Matsumura, Tomoko Kojidani, Yuji Kamioka, Seiichi Uchida, Tokuko Haraguchi, Akatsuki Kimura, Fumiko Toyoshima, Interphase adhesion geometry is transmitted to an internal regulator for spindle orientation via caveolin-1, Nature Communications, 2016.06.
83. Liuan Wang, Seiichi Uchida, Wei Fan, Jun Sun, Globally Optimal Text Line Extraction Based on K-Shortest Paths Algorithm, 12th IAPR International Workshop on Document Analysis Systems, DAS 2016 Proceedings - 12th IAPR International Workshop on Document Analysis Systems, DAS 2016, 10.1109/DAS.2016.12, 335-339, 2016.06, The task of text line extraction in images is a crucial prerequisite for content-based image understanding applications. In this paper, we propose a novel text line extraction method based on k-shortest paths global optimization in images. Firstly, the candidate connected components are extracted by reformulating it as Maximal Stable Extremal Region (MSER) results in images. Then, the directed graph is built upon the connected component nodes with edges comprising of unary and pairwise cost function. Finally, the text line extraction problem is solved using the k-shortest paths optimization algorithm by taking advantage of the particular structure of the directed graph. Experimental results on public dataset demonstrate the effectiveness of proposed method in comparison with state-of-the-art methods..
84. Shigeru Matsumura, Tomoko Kojidani, Yuji Kamioka, Seiichi Uchida, Tokuko Haraguchi, Akatsuki Kimura, Fumiko Toyoshima, Interphase adhesion geometry is transmitted to an internal regulator for spindle orientation via caveolin-1, Nature communications, 10.1038/ncomms11858, 7, 2016.06, Despite theoretical and physical studies implying that cell-extracellular matrix adhesion geometry governs the orientation of the cell division axis, the molecular mechanisms that translate interphase adhesion geometry to the mitotic spindle orientation remain elusive. Here, we show that the cellular edge retraction during mitotic cell rounding correlates with the spindle axis. At the onset of mitotic cell rounding, caveolin-1 is targeted to the retracting cortical region at the proximal end of retraction fibres, where ganglioside GM1-enriched membrane domains with clusters of caveola-like structures are formed in an integrin and RhoA-dependent manner. Furthermore, Gαi1-LGN-NuMA, a well-known regulatory complex of spindle orientation, is targeted to the caveolin-1-enriched cortical region to guide the spindle axis towards the cellular edge retraction. We propose that retraction-induced cortical heterogeneity of caveolin-1 during mitotic cell rounding sets the spindle orientation in the context of adhesion geometry..
85. Markus Goldstein, Seiichi Uchida, A Comparative Evaluation of Unsupervised Anomaly Detection Algorithms for Multivariate Data, PLoS ONE, 10.1371/journal.pone.0152173, 11, 4, e0152173, 2016.04.
86. Liuan Wang, Seiichi Uchida, Wei Fan, Jun Sun, Globally Optimal Text Line Extraction based on K-Shortest Paths algorithm, Proceedings of The 12th IAPR International Workshop on Document Analysis Systems (DAS2016), 2016.03.
87. Kana Aoki, Fumiyo Maeda, Tomoya Nagasako, Seiichi Uchida, Junichi Ikenouchi, A RhoA and Rnd3 cycle regulates actin reassembly during membrane blebbing, Proceedings of the National Academy of Sciences (PNAS), 2016.03.
88. Kana Aoki, Fumiyo Maeda, Tomoya Nagasako, Yuki Mochizuki, Seiichi Uchida, Junichi Ikenouchi, A RhoA and Rnd3 cycle regulates actin reassembly during membrane blebbing, Proceedings of the National Academy of Sciences of the United States of America, 10.1073/pnas.1600968113, 113, 13, E1863-E1871, 2016.03, The actin cytoskeleton usually lies beneath the plasma membrane. When the membrane-associated actin cytoskeleton is transiently disrupted or the intracellular pressure is increased, the plasma membrane detaches from the cortex and protrudes. Such protruded membrane regions are called blebs. However, the molecular mechanisms underlying membrane blebbing are poorly understood. This study revealed that epidermal growth factor receptor kinase substrate 8 (Eps8) and ezrin are important regulators of rapid actin reassembly for the initiation and retraction of protruded blebs. Live-cell imaging of membrane blebbing revealed that local reassembly of actin filaments occurred at Eps8- and activated ezrin-positive foci of membrane blebs. Furthermore, we found that a RhoA-ROCK-Rnd3 feedback loop determined the local reassembly sites of the actin cortex during membrane blebbing..
89. Markus Goldstein, Seiichi Uchida, A Comparative Study on Outlier Removal from a Large-scale Dataset using Unsupervised Anomaly Detection, Proceedings of The 5th International Conference on Pattern Recognition Applications and Methods (ICPRAM2016), 2016.02.
90. Jiamin Xu, Palaiahnakote Shivakumara, Tong Lu, Chew Lim Tan, 内田誠一, A New Method for Multi Oriented Graphics-Scene-3D Text Classification in Video, Pattern Recognition, 10.1016/j.patcog.2015.07.002, 49, 1, 19-42, 2016.01.
91. Seiichi Uchida, Yuto Shinahara, What does scene text tell us?, 23rd International Conference on Pattern Recognition, ICPR 2016 2016 23rd International Conference on Pattern Recognition, ICPR 2016, 10.1109/ICPR.2016.7900267, 4047-4052, 2016.01, Scene text is one of the most important information sources for our daily life because it has particular functions such as disambiguation and navigation. In contrast, ordinary document text has no such function. Consequently, it is natural to have a hypothesis that scene text and document text have different characteristics. This paper tries to prove this hypothesis by semantic analysis of texts by word2vec, which is a neural network model to give a vector representation of each word. By the vector representation, we can have the semantic distributions of scene text and document text in Euclidean space and then determine their semantic categories by simple clustering. Experimental study reveals several differences between scene text and document text. For example, it is found that scene text is a semantic subset of document text and several semantic categories are very specific to scene text..
92. Markus Goldstein, Seiichi Uchida, A comparative study on outlier removal from a large-scale dataset using unsupervised anomaly detection, 5th International Conference on Pattern Recognition Applications and Methods, ICPRAM 2016 ICPRAM 2016 - Proceedings of the 5th International Conference on Pattern Recognition Applications and Methods, 263-269, 2016, Outlier removal from training data is a classical problem in pattern recognition. Nowadays, this problem becomes more important for large-scale datasets by the following two reasons: First, we will have a higher risk of "unexpected" outliers, such as mislabeled training data. Second, a large-scale dataset makes it more difficult to grasp the distribution of outliers. On the other hand, many unsupervised anomaly detection methods have been proposed, which can be also used for outlier removal. In this paper, we present a comparative study of nine different anomaly detection methods in the scenario of outlier removal from a large-scale dataset. For accurate performance observation, we need to use a simple and describable recognition procedure and thus utilize a nearest neighbor-based classifier. As an adequate large-scale dataset, we prepared a handwritten digit dataset comprising of more than 800,000 manually labeled samples. With a data dimensionality of 16×16=256, it is ensured that each digit class has at least 100 times more instances than data dimensionality. The experimental results show that the common understanding that outlier removal improves classification performance on small datasets is not true for high-dimensional large-scale datasets. Additionally, it was found that local anomaly detection algorithms perform better on this data than their global equivalents..
93. Volkmar Frinken, Seiichi Uchida, Deep BLSTM neural networks for unconstrained continuous handwritten text recognition, 13th International Conference on Document Analysis and Recognition, ICDAR 2015 13th IAPR International Conference on Document Analysis and Recognition, ICDAR 2015 - Conference Proceedings, 10.1109/ICDAR.2015.7333894, 911-915, 2015.11, Recently, two different trends in neural network-based machine learning could be observed. The first one are the introduction of Bidirectional Long Short-Term Memory (BLSTM) neural networks (NN) which made sequences with long-distant dependencies amenable for neural network-based processing. The second one are deep learning techniques, which greatly increased the performance of neural networks, by making use of many hidden layers. In this paper, we propose to combine these two ideas for the task of unconstrained handwriting recognition. Extensive experimental evaluation on the IAM database demonstrate an increase of the recognition performance when using deep learning approaches over commonly used BLSTM neural networks, as well as insight into how different types of hidden layers affect the recognition accuracy..
94. Hiroaki Takebe, Yusuke Uehara, Seiichi Uchida, Efficient anchor graph hashing with data-dependent anchor selection, IEICE Transactions on Information and Systems, 10.1587/transinf.2015EDL8060, E98D, 11, 2030-2033, 2015.11, Anchor graph hashing (AGH) is a promising hashing method for nearest neighbor (NN) search. AGH realizes efficient search by generating and utilizing a small number of points that are called anchors. In this paper, we propose a method for improving AGH, which considers data distribution in a similarity space and selects suitable anchors by performing principal component analysis (PCA) in the similarity space..
95. Seiichi Uchida, Yuji Egashira, Kota Sato, Exploring the world of fonts for discovering the most standard fonts and the missing fonts, 13th International Conference on Document Analysis and Recognition, ICDAR 2015 13th IAPR International Conference on Document Analysis and Recognition, ICDAR 2015 - Conference Proceedings, 10.1109/ICDAR.2015.7333800, 441-445, 2015.11, This paper has two contributions toward understanding the principles in font design. The first contribution of this paper is to discover the most standard font shape of each letter class by analyzing thousands of different fonts. For this analysis, two different methods are used. The first method is congealing for aligning multiple images based on a nonlinear geometric transformation model. The average of the aligned image is considered as a standard font shape. The second method is network analysis for representing font variations as a large-scale relative neighborhood graph (RNG) and then finding its center. The font corresponding to the center is considered as the standard font shape. Both of the standard font shapes given by the two methods are plain without decoration, serif, or slant, and thus give an objective reason why we consider the plain font as the typical font shape. The second contribution is to utilize the RNG and the pairwise congealing technique for discovering unexplored font designs and then generating totally new fonts automatically..
96. Dimosthenis Karatzas, Lluis Gomez-Bigorda, Anguelos Nicolaou, Suman Ghosh, Andrew Bagdanov, Masakazu Iwamura, Jiri Matas, Lukas Neumann, Vijay Ramaseshan Chandrasekhar, Shijian Lu, Faisal Shafait, Seiichi Uchida, Ernest Valveny, ICDAR 2015 competition on Robust Reading, 13th International Conference on Document Analysis and Recognition, ICDAR 2015 13th IAPR International Conference on Document Analysis and Recognition, ICDAR 2015 - Conference Proceedings, 10.1109/ICDAR.2015.7333942, 1156-1160, 2015.11, Results of the ICDAR 2015 Robust Reading Competition are presented. A new Challenge 4 on Incidental Scene Text has been added to the Challenges on Born-Digital Images, Focused Scene Images and Video Text. Challenge 4 is run on a newly acquired dataset of 1,670 images evaluating Text Localisation, Word Recognition and End-to-End pipelines. In addition, the dataset for Challenge 3 on Video Text has been substantially updated with more video sequences and more accurate ground truth data. Finally, tasks assessing End-to-End system performance have been introduced to all Challenges. The competition took place in the first quarter of 2015, and received a total of 44 submissions. Only the tasks newly introduced in 2015 are reported on. The datasets, the ground truth specification and the evaluation protocols are presented together with the results and a brief summary of the participating methods..
97. Ryosuke Kakisako, Seiichi Uchida, Frinken Volkmar, Learning non-Markovian constraints for handwriting recognition, 13th International Conference on Document Analysis and Recognition, ICDAR 2015 13th IAPR International Conference on Document Analysis and Recognition, ICDAR 2015 - Conference Proceedings, 10.1109/ICDAR.2015.7333801, 2015-November, 446-450, 2015.11, Recently, the horizon of dynamic time warping (DTW) for matching two sequential patterns has been extended to deal with non-Markovian constraints. The non-Markovian constraints regulate the matching in a wider scale, whereas Markovian constraints regulate the matching only locally. The global optimization of the non-Markovian DTW is proved to be solvable in polynomial time by a graph cut algorithm. The main contribution of this paper is to reveal what is the best constraint for handwriting recognition by using the non-Markovian DTW. The result showed that the best constraint is not a Markovian but a totally non-Markovian constraint that regulates the matching between very distant points; that is, it was proved that the conventional Markovian DTW has a clear limitation and the non- Markovian DTW should be more focused in future research..
98. Masanori Goto, Ryosuke Ishida, Seiichi Uchida, Preselection of support vector candidates by relative neighborhood graph for large-scale character recognition, 13th International Conference on Document Analysis and Recognition, ICDAR 2015 13th IAPR International Conference on Document Analysis and Recognition, ICDAR 2015 - Conference Proceedings, 10.1109/ICDAR.2015.7333773, 306-310, 2015.11, We propose a pre-selection method for training support vector machines (SVM) with a large-scale dataset. Specifically, the proposed method selects patterns around the class boundary and the selected data is fed to train an SVM. For the selection, that is, searching for boundary patterns, we utilize a relative neighborhood graph (RNG). An RNG has an edge for each pair of neighboring patterns and thus, we can find boundary patterns by looking for edges connecting patterns from different classes. Through large-scale handwritten digit pattern recognition experiments, we show that the proposed pre-selection method accelerates SVM training process 5-15 times faster without degrading recognition accuracy..
99. D. Barbuzzi, G. Pirlo, S. Uchida, V. Frinken, D. Impedovo, Similarity-based regularization for semi-supervised learning for handwritten digit recognition, 13th International Conference on Document Analysis and Recognition, ICDAR 2015 13th IAPR International Conference on Document Analysis and Recognition, ICDAR 2015 - Conference Proceedings, 10.1109/ICDAR.2015.7333734, 101-105, 2015.11, This paper presents an experimental analysis on the use of semi-supervised learning in the handwritten digit recognition field. More specifically, two new feedback-based techniques for retraining individual classifiers in a multi-expert scenario are discussed. These new methods analyze the final decision provided by the multi-expert system so that sample classified with a confidence greater than a specific threshold is used to update the system itself. Experimental results carried out on the CEDAR (handwritten digits) database are presented. In particular, error rate, similarity index and a new correlation score among them are considered in order to evaluate the best retraining rule. For the experimental evaluation, an SVM classifier and five different combination techniques at abstract and measurement level have been used. Finally, the results show that iterating the feedback process, on different multi-expert systems built with the five combination techniques, one retraining rule is winning over the other respect to the best correlation score..
100. Brian Iwana, Seiichi Uchida, Kaspar Riesen, Volkmar Frinken, Tackling temporal pattern recognition by vector space embedding, 13th International Conference on Document Analysis and Recognition, ICDAR 2015 13th IAPR International Conference on Document Analysis and Recognition, ICDAR 2015 - Conference Proceedings, 10.1109/ICDAR.2015.7333875, 816-820, 2015.11, This paper introduces a novel method of reducing the number of prototype patterns necessary for accurate recognition of temporal patterns. The nearest neighbor (NN) method is an effective tool in pattern recognition, but the downside is it can be computationally costly when using large quantities of data. To solve this problem, we propose a method of representing the temporal patterns by embedding dynamic time warping (DTW) distance based dissimilarities in vector space. Adaptive boosting (AdaBoost) is then applied for classifier training and feature selection to reduce the number of prototype patterns required for accurate recognition. With a data set of handwritten digits provided by the International Unipen Foundation (iUF), we successfully show that a large quantity of temporal data can be efficiently classified produce similar results to the established NN method while performing at a much smaller cost..
101. Renwu Gao, Shoma Eguchi, Seiichi Uchida, True color distributions of scene text and background, 13th International Conference on Document Analysis and Recognition, ICDAR 2015 13th IAPR International Conference on Document Analysis and Recognition, ICDAR 2015 - Conference Proceedings, 10.1109/ICDAR.2015.7333813, 2015-November, 506-510, 2015.11, Color feature, as one of the low level features, plays important role in image processing, object recognition and other fields. For example, in the task of scene text detection and recognition, lots of methodologies employ features that utilize color contrast of text and the corresponding background for connected component extraction. However, the true distributions of text and its background, in terms of color, is still not examined because it requires an enough number of scene text database with pixel-level labelled text/non-text ground truth. To clarify the relationship between text and its background, in this paper, we aim at investigating the color non-parametric distribution of text and its background using a large database that contains 3018 scene images and 98,600 characters. The results of our experiments show that text and its background can be discriminated by means of color, therefore color feature can be used for scene text detection..
102. Seiichi Uchida, Yuji Egashira, Kota Sato, Exploring the World of Fonts for Discovering the Most Standard Fonts and the Missing Fonts, Proceedings of The 13th International Conference on Document Analysis and Recognition (ICDAR 2015, Nancy, France), 2015.08.
103. Ryosuke Kakisako, Seiichi Uchida, Volkmar Frinken, Learning Non-Markovian Constraints for Handwriting Recognition, Proceedings of The 13th International Conference on Document Analysis and Recognition (ICDAR 2015, Nancy, France), 2015.08.
104. Volkmar Frinken, Seiichi Uchida, Deep BLSTM Neural Networks for Unconstrained Continuous Handwritten Text Recognition, Proceedings of The 13th International Conference on Document Analysis and Recognition (ICDAR 2015, Nancy, France), 2015.08.
105. Masanori Goto, Ryosuke Ishida, Seiichi Uchida, Preselection of Support Vector Candidates by Relative Neighborhood Graph for Large-Scale Character Recognition, Proceedings of The 13th International Conference on Document Analysis and Recognition (ICDAR 2015, Nancy, France), 2015.08.
106. Dimosthenis Karatzas, Lluis Gomez-Bigorda, Anguelos Nicolaou, Suman Ghosh, Andrew Bagdanov, Masakazu Iwamura, Jiri Matas, Lukas Neumann, Vijay Ramaseshan Chandrasekhar, Shijian Lu, Faisal Shafait, Seiichi Uchida, Ernest Valveny, ICDAR 2015 Competition on Robust Reading, Proceedings of The 13th International Conference on Document Analysis and Recognition (ICDAR 2015, Nancy, France), 2015.08.
107. Renwu Gao, Shoma Eguchi, Seiichi Uchida, True Color Distributions of Scene Text and Background, Proceedings of The 13th International Conference on Document Analysis and Recognition (ICDAR 2015, Nancy, France), 2015.08.
108. Brian Iwana, Seiichi Uchida, Kaspar Riesen, Volkmar Frinken, Tackling Pattern Recognition by Vector Space Embedding, Proceedings of The 13th International Conference on Document Analysis and Recognition (ICDAR 2015, Nancy, France), 2015.08.
109. Donato Barbuzzi, Giuseppe Pirlo, Seiichi Uchida, Volkmar Frinken, Donato Impedovo, Similarity-based Regularization for Semi-Supervised Learning for Handwritten Digit Recognition, Proceedings of The 13th International Conference on Document Analysis and Recognition (ICDAR 2015, Nancy, France), 2015.08.
110. Seiichi Uchida, Image-informatics for biology and biology for image-informatics, Journal of the Institute of Electronics, Information and Communication Engineers, 98, 7, 597-603, 2015.07.
111. Faisal Shafait, Dimosthenis Karatzas, Seiichi Uchida, Masakazu Iwamura, Preface, International Journal on Document Analysis and Recognition, 10.1007/s10032-015-0244-0, 18, 2, 109-110, 2015.06.
112. Andreas Fischer, Seiichi Uchida, Volkmar Frinken, Kaspar Riesen, Horst Bunke, Improving Hausdorff Edit Distance Using Structural Node Context, Proceedings of the The 10th IAPR-TC15 Workshop on Graph-based Representations in Pattern Recognition, 2015.05.
113. Koichi Kise, Masakazu Iwamura, Shin Ichiro Omachi, Seiichi Uchida, A trial for development of fundamental technologies for new usage of character and document media, Journal of the Institute of Electronics, Information and Communication Engineers, 98, 4, 311-327, 2015.04.
114. Koichi Kise, Shinichiro Omachi, Seiichi Uchida, Masakazu Iwamura, Marcus Liwicki, Data Embedding into Characters, IEICE Transactions on Information & Systems, E98-D, 1.0, 2015.01.
115. Koichi Kise, Shinichiro Omachi, Seiichi Uchida, Masakazu Iwamura, Marcus Liwicki, Data embedding into characters, IEICE Transactions on Information and Systems, 10.1587/transinf.2014MUI0002, E98D, 1, 10-20, 2015.01, This paper reviews several trials of re-designing conventional communication medium, i.e., characters, for enriching their functions by using data-embedding techniques. For example, characters are redesigned to have better machine-readability even under various geometric distortions by embedding a geometric invariant into each character image to represent class label of the character. Another example is to embed various information into handwriting trajectory by using a new pen device, called a data-embedding pen. An experimental result showed that we can embed 32-bit information into a handwritten line of 5 cm length by using the pen device. In addition to those applications, we also discuss the relationship between data-embedding and pattern recognition in a theoretical point of view. Several theories tell that if we have appropriate supplementary information by data-embedding, we can enhance pattern recognition performance up to 100%..
116. Andreas Fischer, Seiichi Uchida, Volkmar Frinken, Kaspar Riesen, Horst Bunke, Improving Hausdorff edit distance using structural node context, 10th IAPR-TC-15 International Workshop on Graph-Based Representations in Pattern Recognition, GbRPR 2015 Graph-Based Representations in Pattern Recognition - 10th IAPR-TC-15 InternationalWorkshop, GbRPR 2015, Proceedings, 10.1007/978-3-319-18224-7_15, 148-157, 2015.01, In order to cope with the exponential time complexity of graph edit distance, several polynomial-time approximation algorithms have been proposed in recent years. The Hausdorff edit distance is a quadratic-time matching procedure for labeled graphs which reduces the edit distance to a correspondence problem between local substructures. In its original formulation, nodes and their adjacent edges have been considered as local substructures. In this paper, we integrate a more general structural node context into the matching procedure based on hierarchical subgraphs. In an experimental evaluation on diverse graph data sets, we demonstrate that the proposed generalization of Hausdorff edit distance can significantly improve the accuracy of graph classification while maintaining low computational complexity..
117. Renwu Gao, Seiichi Uchida, Asif Shahab, Faisal Shafait, Volkmar Frinken, Visual Saliency Models for Text Detection in Real World, PLoS ONE, 9.0, 12.0, 2014.12.
118. Volkmar Frinken, Ryosuke Kakisako, Seiichi Uchida, A Novel HMM Decoding Algorithm Permitting Long-Term Dependencies and Its Application to Handwritten Word Recognition, 14th International Conference on Frontiers in Handwriting Recognition, ICFHR 2014 Proceedings - 14th International Conference on Frontiers in Handwriting Recognition, ICFHR 2014, 10.1109/ICFHR.2014.29, 128-133, 2014.12, A new decoding for hidden Markov models is presented. As opposed to the commonly used Viterbi algorithm, it is based on the Min-Cut/Max-Flow algorithm instead of dynamic programming. Therefore non-Markovian long-term dependencies can easily be added to influence the decoding path while still finding the optimal decoding in polynomial time. We demonstrate through an experimental evaluation how these constraints can be used to improve an HMM-based handwritten word recognition system that model words via linear character-HMM by restricting the length of each character..
119. Muhammad Imran Malik, Marcus Liwicki, Andreas Dengel, Seiichi Uchida, Volkmar Frinken, Automatic Signature Stability Analysis and Verification Using Local Features, 14th International Conference on Frontiers in Handwriting Recognition, ICFHR 2014 Proceedings - 14th International Conference on Frontiers in Handwriting Recognition, ICFHR 2014, 10.1109/ICFHR.2014.109, 621-626, 2014.12, The purpose of writing this paper is two-fold. First, it presents a novel signature stability analysis based on signature's local / part-based features. The Speeded Up Local features (SURF) are used for local analysis which give various clues about the potential areas from whom the features should be exclusively considered while performing signature verification. Second, based on the results of the local stability analysis we present a novel signature verification system and evaluate this system on the publicly available dataset of forensic signature verification competition, 4NSigComp2010, which contains genuine, forged, and disguised signatures. The proposed system achieved an EER of 15%, which is considerably very low when compared against all the participants of the said competition. Furthermore, we also compare the proposed system with some of the earlier reported systems on the said data. The proposed system also outperforms these systems..
120. Volkmar Frinken, Yutaro Iwakiri, Ryosuke Ishida, Kensho Fujisaki, Seiichi Uchida, Improving point of view scene recognition by considering textual data, 22nd International Conference on Pattern Recognition, ICPR 2014 Proceedings - International Conference on Pattern Recognition, 10.1109/ICPR.2014.512, 2966-2971, 2014.12, At the current rate of technological advancement and social acceptance thereof, it will not be long before wearable devices will be common that constantly record the field of view of the user. We introduce a new database of image sequences, taken with a first person view camera, of realistic, everyday scenes. As a distinguishing feature, we manually transcribed the scene text of each image. This way, sophisticated OCR algorithms can be simulated that can help in the recognition of the location and the activity. To test this hypothesis, we performed a set of experiments using visual features, textual features, and a combination of both. We demonstrate that, although not very powerful when considered alone, the textual information improves the overall recognition rates..
121. Kohei Inai, Mårten Pålsson, Volkmar Frinken, Yaokai Feng, Seiichi Uchida, Selective concealment of characters for privacy protection, 22nd International Conference on Pattern Recognition, ICPR 2014 Proceedings - International Conference on Pattern Recognition, 10.1109/ICPR.2014.66, 333-338, 2014.12, A method of concealing characters is proposed for degrading legibility of privacy sensitive textual information in natural scene, such as car license plate numbers and name tags. An important property of the proposed method is that it realizes selective concealing of characters, that is, the proposed method degrades legibility of character regions without degrading the quality of non-character regions. This selective concealment is realized because characters have special concealment characteristics. Specifically, character legibility can be degraded by damaging stroke structure by using exemplar-based image in painting, which does not affect non-character regions. Experimental results of qualitative and quantitative evaluations have proven that the selective concealment is practically possible. Furthermore, the quantitative evaluation through a subjective experiment revealed appropriate setups of image in painting for maximizing selective concealment performance..
122. Renwu Gao, Seiichi Uchida, Asif Shahab, Faisal Shafait, Volkmar Frinken, Visual saliency models for text detection in real world, PloS one, 10.1371/journal.pone.0114539, 9, 12, 2014.12, This paper evaluates the degree of saliency of texts in natural scenes using visual saliency models. A large scale scene image database with pixel level ground truth is created for this purpose. Using this scene image database and five state-of-theart models, visual saliency maps that represent the degree of saliency of the objects are calculated. The receiver operating characteristic curve is employed in order to evaluate the saliency of scene texts, which is calculated by visual saliency models. A visualization of the distribution of scene texts and non-texts in the space constructed by three kinds of saliency maps, which are calculated using Itti's visual saliency model with intensity, color and orientation features, is given. This visualization of distribution indicates that text characters are more salient than their non-text neighbors, and can be captured from the background. Therefore, scene texts can be extracted from the scene images. With this in mind, a new visual saliency architecture, named hierarchical visual saliency model, is proposed. Hierarchical visual saliency model is based on Itti's model and consists of two stages. In the first stage, Itti's model is used to calculate the saliency map, and Otsu's global thresholding algorithm is applied to extract the salient region that we are interested in. In the second stage, Itti's model is applied to the salient region to calculate the final saliency map. An experimental evaluation demonstrates that the proposed model outperforms Itti's model in terms of captured scene texts..
123. Kyoko Chiba, Masahiko Araseki, Keisuke Nozawa, Keiko Furukori, Yoichi Araki, Takahide Matsushima, Tadashi Nakaya,Saori Hata, Yuhki Saito, Seiichi Uchida, Yasushi Okada, Angus C. Nairn, Roger J. Davis, Tohru Yamamoto, Masataka Kinjo, Hidenori Taru, and Toshiharu Suzuki, Quantitative Analysis of APP Axonal Transport in Neurons -- Role of JIP1 in Enhanced APP Anterograde Transport --, Molecular Biology of the Cell, 25.0, 22.0, 2014.11.
124. Kyoko Chiba, Masahiko Araseki, Keisuke Nozawa, Keiko Furukori, Yoichi Araki, Takahide Matsushima, Tadashi Nakaya, Saori Hata, Yuhki Saito, Seiichi Uchida, Yasushi Okada, Angus C. Nairn, Roger J. Davis, Tohru Yamamoto, Masataka Kinjo, Hidenori Taru, Toshiharu Suzuki, Quantitative analysis of APP axonal transport in neurons
Role of JIP1 in enhanced APP anterograde transport, Molecular biology of the cell, 10.1091/mbc.E14-06-1111, 25, 22, 3569-3580, 2014.11, Alzheimer's β-amyloid precursor protein (APP) associates with kinesin-1 via JNK-interacting protein 1 (JIP1); however, the role of JIP1 in APP transport by kinesin-1 in neurons remains unclear. We performed a quantitative analysis to understand the role of JIP1 in APP axonal transport. In JIP1-deficient neurons, we find that both the fast velocity (∼2.7 μm/s) and high frequency (66%) of anterograde transport of APP cargo are impaired to a reduced velocity (∼1.83 μm/s) and a lower frequency (45%). We identified two novel elements linked to JIP1 function, located in the central region of JIP1b, that interact with the coiled-coil domain of kinesin light chain 1 (KLC1), in addition to the conventional interaction of the JIP1b 11-amino acid C-terminal (C11) region with the tetratricopeptide repeat of KLC1. High frequency of APP anterograde transport is dependent on one of the novel elements in JIP1b. Fast velocity of APP cargo transport requires the C11 domain, which is regulated by the second novel region of JIP1b. Furthermore, efficient APP axonal transport is not influenced by phosphorylation of APP at Thr-668, a site known to be phosphorylated by JNK. Our quantitative analysis indicates that enhanced fast-velocity and efficient high-frequency APP anterograde transport observed in neurons are mediated by novel roles of JIP1b..
125. Cai Wenjie, Seiichi Uchida, Hiroaki Sakoe, Methods for Stroke-Order Free Online Multi-Stroke Character Recognition, Frontiers of Computer Science, 8.0, 5.0, 2014.10.
126. Wenjie Cai, Seiichi Uchida, Hiroaki Sakoe, Comparative performance analysis of stroke correspondence search methods for stroke-order free online multi-stroke character recognition, Frontiers of Computer Science, 10.1007/s11704-014-3207-6, 8, 5, 773-784, 2014.10, For stroke-order free online multi-stroke character recognition, stroke-to-stroke correspondence search between an input pattern and a reference pattern plays an important role to deal with the stroke-order variation. Although various methods of stroke correspondence have been proposed, no comparative study for clarifying the relative superiority of those methods has been done before. In this paper, we firstly review the approaches for solving the stroke-order variation problem. Then, five representative methods of stroke correspondence proposed by different groups, including cube search (CS), bipartite weighted matching (BWM), individual correspondence decision (ICD), stable marriage (SM), and deviation-expansion model (DE), are experimentally compared, mainly in regard of recognition accuracy and efficiency. The experimental results on an online Kanji character dataset, showed that 99.17%, 99.17%, 96.37%, 98.54%, and 96.59% were attained by CS, BWM, ICD, SM, and DE, respectively as their recognition rates. Extensive discussions are made on their relative superiorities and practicalities..
127. Rong Huang, Kyung hyune Rhee and Seichi Uchida, A Parallel Image Encryption Method Based on Compressive Sensing, Multimedia Tools and Applications, 72.0, 1.0, 2014.09.
128. Ryota Ogata, Minoru Mori, Volkmar Frinken and Seiichi Uchida, Constrained AdaBoost for Totally-Ordered Global Features, Proceedings of the 14th International Conference on Frontiers in Handwriting Recognition, 2014.09.
129. Volkmar Frinken, Ryosuke Kakisako and Seiichi Uchida, A Novel HMM Decoding Algorithm Permitting Long-Term Dependencies and its Application to Handwritten Word Recognition, Proceedings of the 14th International Conference on Frontiers in Handwriting Recognition, 2014.09.
130. Muhammad Imran Malik, Marcus Liwicki, Andreas Dengel, Seiichi Uchida and Volkmar Frinken, Automatic Signatures Stability Analysis And Verification Using Local Features, Proceedings of the 14th International Conference on Frontiers in Handwriting Recognition, 2014.09.
131. R. Huang, K. H. Rhee, S. Uchida, A parallel image encryption method based on compressive sensing, Multimedia Tools and Applications, 10.1007/s11042-012-1337-0, 72, 1, 71-93, 2014.09, Recently, compressive sensing-based encryption methods which combine sampling, compression and encryption together have been proposed. However, since the quantized measurement data obtained from linear dimension reduction projection directly serve as the encrypted image, the existing compressive sensing-based encryption methods fail to resist against the chosen-plaintext attack. To enhance the security, a block cipher structure consisting of scrambling, mixing, S-box and chaotic lattice XOR is designed to further encrypt the quantized measurement data. In particular, the proposed method works efficiently in the parallel computing environment. Moreover, a communication unit exchanges data among the multiple processors without collision. This collision-free property is equivalent to optimal diffusion. The experimental results demonstrate that the proposed encryption method not only achieves the remarkable confusion, diffusion and sensitivity but also outperforms the existing parallel image encryption methods with respect to the compressibility and the encryption speed..
132. Volkmar Frinken, Yutaro Iwakiri, Ryosuke Ishida, Kensho Fujisaki, Seiichi Uchida, Improving Point of View Scene Recognition by Considering Textual Data, Proceedings of the 22nd International Conference on Pattern Recognition, 2014.08.
133. Kohei Inai, Marten Palsson, Volkmar Frinken, Yaokai Feng, Seiichi Uchida, Selective Concealment of Characters for Privacy Protection, Proceedings of the 22nd International Conference on Pattern Recognition, 2014.08.
134. Markus Weber, Christopher Scholzel, Marcus Liwicki, Seiichi Uchida, Didier Stricker, LSTM-Based Early Recognition of Motion Patterns, Proceedings of the 22nd International Conference on Pattern Recognition, 2014.08.
135. Volkmar Frinken, Nilanjana Bhattacharya, Seiichi Uchida and Umapada Pal, Improved BLSTM Neural Networks for Recognition of On-line Bangla Complex Words, Proceedings of Joint International Workshops on Statistical Techniques in Pattern Recognition and Structural and Syntactic Pattern Recognition, 2014.08.
136. Kyoko Chiba, Yuki Shimada, Masataka Kinjo, Toshiharu Suzuki, Seiichi Uchida, Simple and Direct Assembly of Kymographs from Movies Using Kymomaker, Traffic, 15.0, 1.0, 1.0-11.0, 2014.01.
137. Megumi Chikano, Koichi Kise, Masakazu Iwamura, Seiichi Uchida, Shinichiro Omachi, Recovery and Localization of Handwritings by a Camera-Pen Based on
Tracking and Document Image Retrieval, Pattern Recognition Letters, 35.0, 1.0, 214.0-224.0, 2014.01.
138. Minoru Mori, Seiichi Uchida, Hitoshi Sakano, Global Feature for Online Character Recognition, Pattern Recognition Letters, 35.0, 1.0, 142.0-148.0, 2014.01.
139. Marcus Liwicki, Seiichi Uchida, Akira Yoshida, Masakazu Iwamura, Shinichiro Omachi, Koichi Kise, More than Ink - Realization of a Data-Embedding Pen, Pattern Recognition Letters, 35.0, 1.0, 246.0-255.0, 2014.01.
140. Renwu Gao, Faisal Shafait, Seiichi Uchida, Yaokai Feng, A hierarchical visual saliency model for character detection in natural scenes, 5th International Workshop on Camera-Based Document Analysis and Recognition, CBDAR 2013 Camera-Based Document Analysis and Recognition - 5th International Workshop, CBDAR 2013, Revised Selected Papers, 10.1007/978-3-319-05167-3_2, 18-29, 2014.01, Visual saliency models have been introduced to the field of character recognition for detecting characters in natural scenes. Researchers believe that characters have different visual properties from their non-character neighbors, which make them salient. With this assumption, characters should response well to computational models of visual saliency. However in some situations, characters belonging to scene text mignt not be as salient as one might expect. For instance, a signboard is usually very salient but the characters on the signboard might not necessarily be so salient globally. In order to analyze this hypothesis in more depth, we first give a view of how much these background regions, such as sign boards, affect the task of saliency-based character detection in natural scenes. Then we propose a hierarchical-saliency method for detecting characters in natural scenes. Experiments on a dataset with over 3,000 images containing scene text show that when using saliency alone for scene text detection, our proposed hierarchical method is able to capture a larger percentage of text pixels as compared to the conventional single-pass algorithm..
141. Ryota Ogata, Minoru Mori, Volkmar Frinken, Seiichi Uchida, Constrained AdaBoost for Totally-Ordered Global Features, 14th International Conference on Frontiers in Handwriting Recognition, ICFHR 2014 Proceedings - 14th International Conference on Frontiers in Handwriting Recognition, ICFHR 2014, 10.1109/ICFHR.2014.72, 393-398, 2014.01, This paper proposes a constrained AdaBoost algorithm for utilizing global features in a dynamic time warping (DTW) framework. Global features are defined as a spatial relationship between temporally-distant points of a temporal pattern and are useful to represent global structure of the pattern. An example is the spatial relationship between the first and the last points of a handwritten pattern of the digit '0'. Those temporally-distant points should be spatially close enough to form a closed circle, whereas those points of '6' should be distant enough. For a temporal pattern of an N-point sequence, it is possible to have N(N - 1)/2 global features. One problem of using the global features is that they are not ordered as a one dimensional sequence any more. Consequently, it is impossible to use them in a left-to-right Markovian model, such as DTW and HMM. The proposed constrained AdaBoost algorithm can select a totally-ordered subset from the set of N(N - 1)/2 global features. Since the totally-ordered features can be arranged as a one-dimensional sequence, they can be incorporated into a DTW framework for compensating nonlinear temporal fluctuation. Since the selection is governed by the AdaBoost framework, the selected features can retain discriminative power..
142. Chihiro Nakamoto, Rong Huang, Sota Koizumi, Ryosuke Ishida, Yaokai Feng, Seiichi Uchida, Font distribution observation by network-based analysis, 5th International Workshop on Camera-Based Document Analysis and Recognition, CBDAR 2013 Camera-Based Document Analysis and Recognition - 5th International Workshop, CBDAR 2013, Revised Selected Papers, 10.1007/978-3-319-05167-3_7, 83-97, 2014.01, The off-the-shelf Optical Character Recognition (OCR) engines return mediocre performance on the decorative characters which usually appear in natural scenes such as signboards. A reasonable way towards the so-called camera-based OCR is to collect a large-scale font set and analyze the distribution of font samples for realizing some character recognition engine which is tolerant to font shape variations. This paper is concerned with the issue of font distribution analysis by network. Minimum Spanning Tree (MST) is employed to construct font network with respect to Chamfer distance. After clustering, some centrality criterion, namely closeness centrality, eccentricity centrality or betweenness centrality, is introduced for extracting typical font samples. The network structure allows us to observe the font shape transition between any two samples, which is useful to create new fonts and recognize unseen decorative characters. Moreover, unlike the Principal Component Analysis (PCA), the font network fulfills distribution visualization through measuring the dissimilarity between samples rather than the lossy processing of dimensionality reduction. Compared with K-means algorithm, network-based clustering has the ability to preserve small size font clusters which generally consist of samples taking special appearances. Experiments demonstrate that the proposed network-based analysis is an effective way to grasp font distribution, and thus provides helpful information for decorative character recognition..
143. Minoru Mori, Seiichi Uchida, Hitoshi Sakano, Global feature for online character recognition, Pattern Recognition Letters, 10.1016/j.patrec.2013.03.036, 35, 1, 142-148, 2014.01, This paper focuses on the importance of global features for online character recognition. Global features represent the relationship between two temporally distant points in a handwriting pattern. For example, it can be defined as the relative vector of two xy-coordinate features of two temporally separated points. Most existing online character recognition methods do not utilize global features, since their non-Markovian property prevents the use of the traditional recognition methodologies, such as dynamic time warping and hidden Markov models. However, we can understand the importance of, for example, the relationship between the starting and the ending points by attempting to discriminate "0" and "6". This relationship cannot be represented by local features defined at individual points but by global features. Since O(N2) global features can be extracted from a handwriting pattern with N points, selecting those that are truly discriminative is very important. In this paper, AdaBoost is employed for feature selection. Experiments prove that many global features are discriminative and the combined use of local and global features can improve the recognition accuracy..
144. Volkmar Frinken, Nilanjana Bhattacharya, Seiichi Uchida, Umapada Pal, Improved BLSTM neural networks for recognition of on-line bangla complex words, Joint IAPR International Workshop on Structural, Syntactic, and Statistical Pattern Recognition, S+SSPR 2014 Structural, Syntactic, and Statistical Pattern Recognition - Joint IAPR International Workshop, S+SSPR 2014, Proceedings, 10.1007/978-3-662-44415-3_41, 404-413, 2014.01, While bi-directional long short-term (BLSTM) neural network have been demonstrated to perform very well for English or Arabic, the huge number of different output classes (characters) encountered in many Asian fonts, poses a severe challenge. In this work we investigate different encoding schemes of Bangla compound characters and compare the recognition accuracies. We propose to model complex characters not as unique symbols, which are represented by individual nodes in the output layer. Instead, we exploit the property of long-distance-dependent classification in BLSTM neural networks. We classify only basic strokes and use special nodes which react to semantic changes in the writing, i.e., distinguishing inter-character spaces from intra-character spaces. We show that our approach outperforms the common approaches to BLSTM neural network-based handwriting recognition considerably..
145. Markus Weber, Marcus Liwicki, Didier Stricker, Christopher Scholzel, Seiichi Uchida, LSTM-based early recognition of motion patterns, 22nd International Conference on Pattern Recognition, ICPR 2014 Proceedings - International Conference on Pattern Recognition, 10.1109/ICPR.2014.611, 3552-3557, 2014.01, In this paper a method for Early Recognition (ER) of Motion Templates (MTs) is presented. We define ER as an algorithm to provide recognition results before a motion sequence is completed. In our experiments we apply Long Short-Term Memory (LSTM) and optimize the training for the task of recognizing the motion template as early as possible. The evaluation has shown that the recognition accuracy for a frame-by-frame classification the LSTM achieves a recognition accuracy of 88% if no training data of the person him/herself is included, and 92% if the training data also contains motion sequences of the person. Furthermore, the average earliness - the number of time frames it takes before the LSTM correctly classifies a motion pattern - is around 24.77 frames, which is less than a second with the used tracking technology, i.e., the Microsoft Kinect..
146. Osamu Matsuda, Noriyuki Suetsugu, Seiichi Uchida, Masamitsu Wada, Koh Iba, Molecular genetic application of hyperspectral image sensing as a method for high-throughput quantitative phenotype analysis, Japanese Journal of Ecology, 64, 3, 205-213, 2014.01.
147. Marcus Liwicki, Seiichi Uchida, Akira Yoshida, Masakazu Iwamura, Shinichiro Omachi, Koichi Kise, More than ink - Realization of a data-embedding pen, Pattern Recognition Letters, 10.1016/j.patrec.2012.09.001, 35, 1, 246-255, 2014.01, In this paper we present a novel digital pen device, called data-embedding pen, for enhancing the value of handwriting on physical paper. This pen produces an additional ink-dot sequence along a written stroke during writing. This ink-dot sequence represents arbitrary information, such as writer's name and writing date. Since the information is placed on the paper as an ink-dot sequence, it can be retrieved just by scanning or photographing the paper. In addition to the hardware of the data-embedding pen, this paper also proposes a coding scheme for reliable data-embedding and retrieval. In fact, the physical data-embedding on a paper will undergo various severe errors and therefore a robust coding scheme is necessary. Through experiments on data written by two writers, we show that we can embed 32 bits on short and simple or even on more complex patterns and finally retrieve them with a high reliability..
148. Megumi Chikano, Koichi Kise, Masakazu Iwamura, Seiichi Uchida, Shinichiro Omachi, Recovery and localization of handwritings by a camera-pen based on tracking and document image retrieval, Pattern Recognition Letters, 10.1016/j.patrec.2012.10.003, 35, 1, 214-224, 2014.01, We propose a camera-based method for digital recovery of handwritings on ordinary paper. Our method is characterized by the following two points: (1) it requires no special device such as special paper other than a camera-pen to recover handwritings, (2) if the handwriting is on a printed document, the method is capable of localizing it onto an electronic equivalent of the printed document. The above points are enabled by the following processing. The handwriting is recovered by the LK tracking to trace the move of the pen-tip. The recovered shape is localized onto the corresponding part of the electronic document with the help of document image retrieval called LLAH (locally likely arrangement hashing). A new framework for stably estimating the homography from a camera-captured image to the corresponding electronic document allows us to localize the recovered handwritings accurately. We experimentally evaluate the accuracy, processing time and memory usage of the proposed method using 30 handwritings. From the comparison to other methods that implement alternative ways for realizing the same functionality, we have confirmed that the proposed method is superior to those other methods..
149. Kyoko Chiba, Yuki Shimada, Masataka Kinjo, Toshiharu Suzuki, Seiichi Uchida, Simple and Direct Assembly of Kymographs from Movies Using KYMOMAKER, Traffic, 10.1111/tra.12127, 15, 1, 1-11, 2014.01, In tracking analysis, the movement of cargos by motor proteins in axons is often represented by a time-space plot termed a 'kymograph'. Manual creation of kymographs is time-consuming and complicated for cell biologists. Therefore, we developed KYMOMAKER, a simple system that automatically creates a kymograph from a movie without generating multiple time-dissected movie stacks. In addition, KYMOMAKER can automatically extract faint vesicle traces, and can thereby effectively analyze cargos expressed at low levels in axons. A filter can be applied to remove traces of non-physiological movements and to extract meaningful traces of anterograde or retrograde cargo transport. For example, only cargos that move at a speed of >0.4μm/second for a distance of >1μm can be included. Another function of KYMOMAKER is to create a color kymograph in which the color of the trace varies according to the position of the fluorescent particle in the axis perpendicular to the long axis of the axon. Such positional information is completely lost in conventional kymographs. KYMOMAKER is an open access program that can be easily used to analyze vesicle transport in axons by cell biologists who do not have specific knowledge of bioimage informatics..
150. Seiichi Uchida, Text localization and recognition in images and video, Handbook of Document Image Processing and Recognition, 10.1007/978-0-85729-859-1_28, 843-883, 2014.01, This chapter reviews techniques on text localization and recognition in scene images captured by camera. Since properties of scene texts are very different from scanned documents in various aspects, specific techniques are necessary to localize and recognize them. In fact, localization of scene text is a difficult and important task because there is no prior information on the location, layout, direction, size, typeface, and color of texts in a scene image in general and there are many textures and patterns similar to characters. In addition, recognition of scene text is also a difficult task because there are many characters distorted by blurring, perspective, nonuniform lighting, and low resolution. Decoration of characters makes the recognition task far more difficult. As reviewed in this chapter, those difficult tasks have been tackled with not only modified versions of conventional OCR techniques but also state-of-the-art computer vision and pattern recognition methodologies..
151. Kensho Fujisaki, Ayumi Hamano, Kenta Aoki, Yaokai Feng, Seiichi Uchida, Masahiko Araseki, Yuki Saito and Toshiharu Suzuki, Detection and Tracking Protein Molecules in Fluorescence Microscopic Video, Proceedings of The 1st International Workshop on BioImage Recognition (BIR'13, Ehime, Japan), 2013.12.
152. Ayumi Hamano, Kensho Fujisaki, Seiichi Uchida and Osamu Shiku, Stable Marriage Algorithm for Tracking Intracellular Objects, Proceedings of The 1st International Workshop on BioImage Recognition (BIR'13, Ehime, Japan), 2013.12.
153. Masanori Goto, Ryosuke Ishida, Yaokai Feng, Seiichi Uchida, Analyzing the distribution of a large-scale character pattern set using relative neighborhood graph, 12th International Conference on Document Analysis and Recognition, ICDAR 2013 Proceedings of the International Conference on Document Analysis and Recognition, ICDAR, 10.1109/ICDAR.2013.10, 3-7, 2013.12, The goal of this research is to understand the true distribution of character patterns. Advances in computer technology for mass storage and digital processing have paved way to process a massive dataset for various pattern recognition problems. If we can represent and analyze the distribution of a large-scale character pattern set directly and understand its relationships deeply, it should be helpful for improving character recognizer. For this purpose, we propose a network analysis method to represent the distribution of patterns using a relative neighborhood graph and its clustered version. In this paper, the properties and validity of the proposed method are confirmed on 410,564 machine-printed digit patterns and 622,660 handwritten digit patterns which were manually ground-truthed and resized to 16 times 16 pixels. Our network analysis method represents the distribution of the patterns without any assumption, approximation or loss..
154. Kensho Fujisaki, Ayumi Hamano, Kenta Aoki, Yaokai Feng, Seiichi Uchida, Masahiko Araseki, Yuki Saito, Toshiharu Suzuki, Detection and tracking protein molecules in fluorescence microscopic video, 2013 1st International Symposium on Computing and Networking, CANDAR 2013 Proceedings - 2013 1st International Symposium on Computing and Networking, CANDAR 2013, 10.1109/CANDAR.2013.47, 270-274, 2013.12, This paper provides a bioimage informatics system of detecting and tracking protein molecules, called APP-GFPs, in a live-cell video captured by a fluorescent microscope. Since both processes encounter many difficulties such as many targets, less appearance information, and heavy background noise, we will try to design the system as robust as possible. Specifically, for the detection, a machine learning-based method is employed. For tracking, a method based on a global optimization strategy is newly developed. Experimental results showed that the speed and direction distributions of molecular motion by the proposed system were very similar to that by manual inspection..
155. Dimosthenis Karatzas, Faisal Shafait, Seiichi Uchida, Masakazu Iwamura, Lluis Gomez I. Bigorda, Sergi Robles Mestre, Joan Mas, David Fernandez Mota, Jon Almazan Almazan, Lluis Pere De Las Heras, ICDAR 2013 robust reading competition, 12th International Conference on Document Analysis and Recognition, ICDAR 2013 Proceedings of the International Conference on Document Analysis and Recognition, ICDAR, 10.1109/ICDAR.2013.221, 1484-1493, 2013.12, This report presents the final results of the ICDAR 2013 Robust Reading Competition. The competition is structured in three Challenges addressing text extraction in different application domains, namely born-digital images, real scene images and real-scene videos. The Challenges are organised around specific tasks covering text localisation, text segmentation and word recognition. The competition took place in the first quarter of 2013, and received a total of 42 submissions over the different tasks offered. This report describes the datasets and ground truth specification, details the performance evaluation protocols used and presents the final results along with a brief summary of the participating methods..
156. Yugo Terada, Rong Huang, Yaokai Feng, Seiichi Uchida, On the possibility of structure learning-based scene character detector, 12th International Conference on Document Analysis and Recognition, ICDAR 2013 Proceedings of the International Conference on Document Analysis and Recognition, ICDAR, 10.1109/ICDAR.2013.101, 472-476, 2013.12, In this paper, we propose a structure learning-based scene character detector which is inspired by the observation that characters have their own inherent structures compared with the background. Graphs are extracted from the thinned binary image to represent the topological line structures of scene contents. Then, a graph classifier, namely gBoost classifier, is trained with the intent to seek out the inherent structures of character and the counterparts of non-character. The experimental results show that the proposed detector achieves the remarkable classification performance with the accuracy of about 70%, which demonstrates the existence and separability of the inherent structures..
157. Song Wang, Seiichi Uchida, Marcus Liwicki, Part-based recognition of arbitrary fonts, 12th International Conference on Document Analysis and Recognition, ICDAR 2013 Proceedings of the International Conference on Document Analysis and Recognition, ICDAR, 10.1109/ICDAR.2013.41, 170-174, 2013.12, In this paper, the part-based recognition method is introduced and applied to the arbitrary font recognition. The principle of the part-based method is to represent the character image as a set of parts and then recognize the image by finding the most possible parts set from the reference database. Since the part-based method does not rely on the global structure of a character, it is supposed to be robust against the variant appearances of the character. The experiment results indicate that it is possible to apply the part-based method to the font recognition, which is always considered as a difficult task by most of the researchers..
158. Rong Huang, Palaiahnakote Shivakumara, Seiichi Uchida, Scene character detection by an edge-ray filter, 12th International Conference on Document Analysis and Recognition, ICDAR 2013 Proceedings of the International Conference on Document Analysis and Recognition, ICDAR, 10.1109/ICDAR.2013.99, 462-466, 2013.12, Edge is a type of valuable clues for scene character detection task. Generally, the existing edge-based methods rely on the assumption of straight text line to prune away the non-character candidates. This paper proposes a new edge-based method, called edge-ray filter, to detect the scene character. The main contribution of the proposed method lies in filtering out complex backgrounds by fully utilizing the essential spatial layout of edges instead of the assumption of straight text line. Edges are extracted by a combination of Canny and Edge Preserving Smoothing Filter (EPSF). To effectively boost the filtering strength of the designed edge-ray filter, we employ a new Edge Quasi-Connectivity Analysis (EQCA) to unify complex edges as well as contour of broken character. Label Histogram Analysis (LHA) then filters out non-character edges and redundant rays through setting proper thresholds. Finally, two frequently-used heuristic rules, namely aspect ratio and occupation, are exploited to wipe off distinct false alarms. In addition to have the ability to handle special scenarios, the proposed method can accommodate dark-on-bright and bright-on-dark characters simultaneously, and provides accurate character segmentation masks. We perform experiments on the benchmark ICDAR 2011 Robust Reading Competition dataset as well as scene images with special scenarios. The experimental results demonstrate the validity of our proposal..
159. Ayumi Hamano, Kensho Fujisaki, Seiichi Uchida, Osamu Shiku, Stable marriage algorithm for tracking intracellular objects, 2013 1st International Symposium on Computing and Networking, CANDAR 2013 Proceedings - 2013 1st International Symposium on Computing and Networking, CANDAR 2013, 10.1109/CANDAR.2013.53, 305-307, 2013.12, Development of automatic multiple intracellular-objects tracking methods is one of the significant challenges in Bioimage-Informatics. The challenge becomes more difficult in case the tracking targets have the same shape and appearance. In order to obtain stable results under that condition, we propose a tracking method based on global optimization. Particularly, we first detect tracking targets by our proposed detection method. Then, we formulate the multiple object tracking problem as a combinatorial optimization problem over a pair of consecutive frames. Finally, we solve the problem by the stable marriage algorithm. In this paper, we describe our proposed detection and tracking methods..
160. Takashi Kimura, Rong Huang, Seiichi Uchida, Masakazu Iwamura, Shinichiro Omachi, Koichi Kise, The reading-life log - Technologies to recognize texts that we read, 12th International Conference on Document Analysis and Recognition, ICDAR 2013 Proceedings of the International Conference on Document Analysis and Recognition, ICDAR, 10.1109/ICDAR.2013.26, 91-95, 2013.12, Reading life log is a type of techniques to automatically and unconsciously record people's reading intentions, interests and habits. Besides, it can also serve as various assistants in our daily life. In this paper, a reading-life log system is implemented by a head-mounted and unobtrusive video camera with a high resolution and a high shutter speed. We utilize DP matching, and propose a text-based frame mosaicing method to integrate multiple frames in a clip. The developed system is tested in the various environments indoor and outdoor. The experimental results show that our system can provide reliable outputs with respect to the most correct responses. The infrequent misregistration between lines also indicates the feasibility and validity of the text-based frame mosaicing..
161. Hiroaki Takebe, Seiichi Uchida, Efficient anchor graph hashing with data-dependent anchor selection, IEICE Transactions on Information & Systems, E96-D, 10.0, 2235.0-2244.0, 2013.10.
162. Rong Huang, Palaiahnakote Shivakumara, Yaokai Feng, Seiichi Uchida, Scene Character Detection and Recognition with Cooperative
Multiple-Hypothesis Framework, IEICE Transactions on Information & Systems, E96-D, 10.0, 2235.0-2244.0, 2013.10.
163. Koichi Ogawara, Masahiro Fukutomi, Seiichi Uchida, Yaokai Feng, A Voting-Based Sequential Pattern Recognition Method, PLoS ONE, 8.0, 10.0, e76980, 2013.10.
164. Koichi Ogawara, Masahiro Fukutomi, Seiichi Uchida, Yaokai Feng, A Voting-Based Sequential Pattern Recognition Method, PloS one, 10.1371/journal.pone.0076980, 8, 10, 2013.10, We propose a novel method for recognizing sequential patterns such as motion trajectory of biological objects (i.e., cells, organelle, protein molecules, etc.), human behavior motion, and meteorological data. In the proposed method, a local classifier is prepared for every point (or timing or frame) and then the whole pattern is recognized by majority voting of the recognition results of the local classifiers. The voting strategy has a strong benefit that even if an input pattern has a very large deviation from a prototype locally at several points, they do not severely influence the recognition result; they are treated just as several incorrect votes and thus will be neglected successfully through the majority voting. For regularizing the recognition result, we introduce partial-dependency to local classifiers. An important point is that this dependency is introduced to not only local classifiers at neighboring point pairs but also to those at distant point pairs. Although, the dependency makes the problem non-Markovian (i.e., higher-order Markovian), it can still be solved efficiently by using a graph cut algorithm with polynomial-order computations. The experimental results revealed that the proposed method can achieve better recognition accuracy while utilizing the above characteristics of the proposed method..
165. Rong Huang, Palaiahnakote Shivakumara, Yaokai Feng, Seiichi Uchida, Scene character detection and recognition with cooperative multiple-hypothesis framework, IEICE Transactions on Information and Systems, 10.1587/transinf.E96.D.2235, E96-D, 10, 2235-2244, 2013.10, To handle the variety of scene characters, we propose a cooperative multiple-hypothesis framework which consists of an image operator set module, an Optical Character Recognition (OCR) module and an integration module. Multiple image operators activated by multiple parameters probe suspected character regions. The OCR module is then applied to each suspected region and returns multiple candidates with weight values for future integration. Without the aid of the heuristic rules which impose constraints on segmentation area, aspect ratio, color consistency, text line orientations, etc., the integration module automatically prunes the redundant detection/recognition and pads the missing detection/recognition. The proposed framework bridges the gap between scene character detection and recognition, in the sense that a practical OCR engine is effectively leveraged for result refinement. In addition, the proposed method achieves the detection and recognition at the character level, which enables dealing with special scenarios such as single character, text along arbitrary orientations or text along curves. We perform experiments on the benchmark ICDAR 2011 Robust Reading Competition dataset which includes a text localization task and a word recognition task. The quantitative results demonstrate that multiple hypotheses outperform a single hypothesis, and be comparable with state-of-the-art methods in terms of recall, precision, F-measure, character recognition rate, total edit distance and word recognition rate. Moreover, two additional experiments are conducted to confirm the simplicity of parameter setting in this proposal..
166. Koichi Kise, Riki Kudo, Masakazu Iwamura, Seiichi Uchida and Shinichiro Omachi, A Proposal of Writing-Life Log and Its Implementation Using a
Retrieval-Based Camera-Pen, Proceedings of the 16th International Graphonomics Society Conference (IGS 2013, Nara, Japan), 86.0-89.0, 2013.08.
167. Wenjie Cai, Seiichi Uchida and Hiroaki Sakoe, An Efficient Radical-Based Algorithm for Stroke-Order Free and
Stroke-Number Free Online Kanji Character Recognition, Proceedings of the 16th International Graphonomics Society Conference (IGS 2013, Nara, Japan), 82.0-85.0, 2013.08.
168. Takafumi Matsuo, Song Wang, Yaokai Feng and Seiichi Uchida, Exploring the Ability of Parts on Recognizing Handwriting Characters, Proceedings of the 16th International Graphonomics Society Conference (IGS 2013, Nara, Japan), 66.0-69.0, 2013.08.
169. Song Wang, Seiichi Uchida and Marcus Liwicki, Part-Based Recognition of Arbitrary Fonts, Proceedings of The 12th International Conference on Document Analysis and Recognition (ICDAR 2013, Washington DC, USA), 170.0-174.0, 2013.08.
170. Masanori Goto, Ryosuke Ishida, Yaokai Feng and Seiichi Uchida, Analyzing the Distribution of a Large-scale Character Pattern Set
Using Relative Neighborhood Graph, Proceedings of The 12th International Conference on Document Analysis and Recognition (ICDAR 2013, Washington DC, USA), 3.0-7.0, 2013.08.
171. Yugo Terada, Rong Huang, Yaokai Feng and Seiichi Uchida, On the Possibility of Structure Learning-Based Scene Character Detector, Proceedings of The 12th International Conference on Document Analysis and Recognition (ICDAR 2013, Washington DC, USA), 472.0-476.0, 2013.08.
172. Takashi Kimura, Rong Huang, Seiichi Uchida, Masakazu Iwamura, Shinichiro Omachi and Koichi Kise, Reading-life Log - Technologies to Recognize Texts That We Read, Proceedings of The 12th International Conference on Document Analysis and Recognition (ICDAR 2013, Washington DC, USA), 91.0-95.0, 2013.08.
173. Rong Huang, Palaiahnakote Shivakumara and Seiichi Uchida, Scene Character Detection by an Edge-Ray Filter, Proceedings of The 12th International Conference on Document Analysis and Recognition (ICDAR 2013, Washington DC, USA), 462.0-466.0, 2013.08.
174. Dimosthenis Karatzas, Faisal Shafait, Seiichi Uchida, Masakazu Iwamura, Lluis Gomez i Bigorda, Sergi Robles Mestre, Joan Mas, David Fernandez Mota, Jon Almazan Almazan, Lluis Pere de las Heras, ICDAR 2013 Robust Reading Competition, Proceedings of The 12th International Conference on Document Analysis and Recognition (ICDAR 2013, Washington DC, USA), 1484.0-1493.0, 2013.08.
175. Renwu Gao, Faisal Shafait, Seiichi Uchida and Yaokai Feng, Saliency inside Saliency - A Hierarchical Usage of Visual Saliency for
Scene Character Detection, Proceedings of The 12th International Conference on Document Analysis and Recognition (ICDAR 2013, Washington DC, USA), 2013.08.
176. Chihiro Nakamoto, Rong Huang, Sota Koizumi, Ryosuke Ishida, Yaokai Feng and Seiichi Uchida, Font Distribution Analysis by Network, Proceedings of The 12th International Conference on Document Analysis and Recognition (ICDAR 2013, Washington DC, USA), 2013.08.
177. Song Wang, Seiichi Uchida, Marcus Liwicki, Yaokai Feng, Part-Based Methods for Handwritten Digit Recognition, Frontiers of Computer Science, 10.1007/s11704-013-2297-x, 7.0, 4.0, 514.0-525.0, 2013.07.
178. Soma Shiraishi, Yaokai Feng, Seiichi Uchida, Skew Estimation by Parts, IEICE Transactions on Information & Systems, 2013.07.
179. Soma Shiraishi, Yaokai Feng, Seiichi Uchida, Skew estimation by parts, IEICE Transactions on Information and Systems, 10.1587/transinf.E96.D.1503, E96-D, 7, 1503-1512, 2013.07, This paper proposes a new part-based approach for skew estimation of document images. The proposed method first estimates skew angles on rather small areas, which are the local parts of characters, and subsequently determines the global skew angle by aggregating those local estimations. A local skew estimation on a part of a skewed character is performed by finding an identical part from prepared upright character images and calculating the angular difference. Specifically, a keypoint detector (e.g. SURF) is used to determine the local parts of characters, and once the parts are described as feature vectors, a nearest neighbor search is conducted in the instance database to identify the parts. Finally, a local skew estimation is acquired by calculating the difference of the dominant angles of brightness gradient of the parts. After the local skew estimation, the global skew angle is estimated by the majority voting of those local estimations, disregarding some noisy estimations. Our experiments have shown that the proposed method is more robust to short and sparse text lines and non-text backgrounds in document images compared to conventional methods..
180. Seiichi Uchida, Image Processing and Recognition for Biological Images, Development Growth and Differentiation, 2013.05.
181. Seiichi Uchida, Image processing and recognition for biological images, Development Growth and Differentiation, 10.1111/dgd.12054, 55, 4, 523-549, 2013.05, This paper reviews image processing and pattern recognition techniques, which will be useful to analyze bioimages. Although this paper does not provide their technical details, it will be possible to grasp their main tasks and typical tools to handle the tasks. Image processing is a large research area to improve the visibility of an input image and acquire some valuable information from it. As the main tasks of image processing, this paper introduces gray-level transformation, binarization, image filtering, image segmentation, visual object tracking, optical flow and image registration. Image pattern recognition is the technique to classify an input image into one of the predefined classes and also has a large research area. This paper overviews its two main modules, that is, feature extraction module and classification module. Throughout the paper, it will be emphasized that bioimage is a very difficult target for even state-of-the-art image processing and pattern recognition techniques due to noises, deformations, etc. This paper is expected to be one tutorial guide to bridge biology and image processing researchers for their further collaboration to tackle such a difficult target. Development, Growth & Differentiation.
182. Chuanjun Liu, Ryohie Yokoyama, Seiichi Uchida, Koji Nakano, Kenshi Hayashi, Odor spatial distribution visualized by a fluorescent imaging sensor, 12th IEEE SENSORS 2013 Conference IEEE SENSORS 2013 - Proceedings, 10.1109/ICSENS.2013.6688508, 2013.01, The detection of spatial-temporal distribution of odors is a significant but challenging work in the field of odor sensors. In this paper a fluorescent imaging sensor system consisting of an excitation light, a fluorescent film and a cooled CCD camera was proposed to visually detect the odor spatial distribution. The visualization principle was based on the fluorescent quenching or enhancing caused by the interaction between the film and odor substances. The change of fluorescent intensity was recorded by the CCD camera and visualized by subtraction image analysis. The proposed sensor could visualize not only the shape and size of an odor flow injected to the film surface, but also the concentration gradient in the interior of odor flows. Additionally, 2D and 3D imaging approaches were used to visualize some special examples related to the spatial distribution of odors, such as the direction of odor flows, the shape of odor marks, and the impact of two different odor flows on the film surface..
183. Rong Huang, Kyung hyune Rhee and Seichi Uchida, A Parallel Image Encryption Method Based on Compressive Sensing, Multimedia Tools and Applications, 2012.12.
184. Seiichi Uchida, Ryosuke Ishida, Akira Yoshida, Wenjie Cai, Yaokai Feng, Character image patterns as big data, 13th International Conference on Frontiers in Handwriting Recognition, ICFHR 2012 Proceedings - 13th International Conference on Frontiers in Handwriting Recognition, ICFHR 2012, 10.1109/ICFHR.2012.190, 479-484, 2012.12, The ambitious goal of this research is to understand the real distribution of character patterns. Ideally, if we can collect all possible character patterns, we can totally understand how they are distributed in the image space. In addition, we also have the perfect character recognizer because we know the correct class for any character image. Of course, it is practically impossible to collect all those patterns - however, if we collect character patterns massively and analyze how the distribution changes according to the increase of patterns, we will be able to estimate the real distribution asymptotically. For this purpose, we use 822,714 manually ground-truthed 32 × 32 handwritten digit patterns in this paper. The distribution of those patterns are observed by nearest neighbor analysis and network analysis, both of which do not make any approximation (such as low-dimensional representation) and thus do not corrupt the details of the distribution..
185. Minoru Mori, Seiichi Uchida, Hitoshi Sakano, Dynamic programming matching with global features for online character recognition, 13th International Conference on Frontiers in Handwriting Recognition, ICFHR 2012 Proceedings - 13th International Conference on Frontiers in Handwriting Recognition, ICFHR 2012, 10.1109/ICFHR.2012.199, 348-353, 2012.12, This paper proposes a dynamic programming (DP) matching method with global features for online character recognition. Many online character recognition methods have utilized the ability of DP matching on compensating temporal fluctuation. On the other hand, DP requires the Markovian property on its matching process. Consequently, most traditional DP matching methods have utilized local information of strokes such as xy-coordinates or local directions as features, because it is easy to satisfy the Markovian property with those features. Unfortunately, these local features cannot represent global structure of character shapes. Although global features that extract global structures of characters have high potential to represent various key characteristics of character shapes, conventional DP matching methods cannot handle global features. This is because the incorporation of global features is not straightforward due to the Markovian property of DP. In this paper we propose a new scheme for DP matching using global features. Our method first selects global features which not only satisfy the Markovian property but also have sufficient discrimination ability. By embedding the selected global features into DP matching process, we can compensate temporal fluctuation while considering the global structure of the pattern. Experimental results show that our methods can enhance the recognition accuracy for online numeral characters..
186. Seiichi Uchida, Masahiro Fukutomi, Koichi Ogawara, Yaokai Feng, Non-Markovian dynamic time warping, 21st International Conference on Pattern Recognition, ICPR 2012 ICPR 2012 - 21st International Conference on Pattern Recognition, 2294-2297, 2012.12, This paper proposes a new dynamic time warping (DTW) method, called non-Markovian DTW. In the conventional DTW, the warping function is optimized generally by dynamic programming (DP) subject to some Markovian constraints which restrict the relationship between neighboring time points. In contrast, the non-Markovian DTW can introduce non-Markovian constraints for dealing with the relationship between points with a large time interval. This new and promising ability of DTW is realized by using graph cut as the optimizer of the warping function instead of DP. Specifically, the conventional DTW problem is first converted as an equivalent minimum cut problem on a graph and then edges representing the non-Markovian constraints are added to the graph. An experiment on online character recognition showed the advantage of using non-Markovian constraints during DTW..
187. Yutaro Iwakiri, Soma Shiraishi, Yaokai Feng, Seiichi Uchida, On the possibility of instance-based stroke recovery, 13th International Conference on Frontiers in Handwriting Recognition, ICFHR 2012 Proceedings - 13th International Conference on Frontiers in Handwriting Recognition, ICFHR 2012, 10.1109/ICFHR.2012.248, 29-34, 2012.12, This paper tackles the stroke recovery problem, which is a typical ill-posed reverse problem, by an instancebased method. The basic idea of the instance-based stroke recovery is to refer to the drawing order of a similar instance. The instance-based method has a strong merit that it can deal with multi-stroke characters and other complex characters without any special consideration. However, it requires a sufficient numbers of instances to cover those various characters. As an initial trial of the instance-based stroke recovery method, this paper describes the principle of the method and then provides several experimental results. The experimental results indicate the potential of the proposed method on recovering the drawing order of complex characters, as expected..
188. Song Wang, Seiichi Uchida, Marcus Liwicki, Part-based method on handwritten texts, 21st International Conference on Pattern Recognition, ICPR 2012 ICPR 2012 - 21st International Conference on Pattern Recognition, 339-342, 2012.12, This paper reports a trial of handwritten text recognition by a part-based method. The part-based method recognizes individual characters by their parts without considering their whole shape. This realizes great robustness to severe deformations. This robustness is also effective for text recognition. Especially, for handwritten texts whose segmentation into individual characters is very difficult by deep touching and heavy slant, the part-based method still can recognize them because it does not request segmentation results to provide their whole shapes. Experimental results using digit sequences proved this robustness..
189. Song Wang, Seiichi Uchida, and Marcus Liwicki, Part-Based Method on Handwritten Texts, Proceedings of the 21st International Conference on Pattern Recognition, 2012.11.
190. Rong Huang, Shinpei Oba, Shivakumara Palaiahnakote, and Seiichi Uchida, Scene Character Detection and Recognition Based on Multiple Hypotheses Framework (PDF), Proceedings of the 21st International Conference on Pattern Recognition, 2012.11.
191. Seiichi Uchida, Masahiro Fukutomi, Koichi Ogawara, and Yaokai Feng, Non-Markovian Dynamic Time Warping, Proceedings of the 21st International Conference on Pattern Recognition, 2012.11.
192. Seiichi Uchida, Satoshi Hokahori, and Yaokai Feng, Analytical Dynamic Programming Matching, Proceedings of the Fifth Workshop on Non-Rigid Shape Analysis and Deformable Image Alignment, 2012.09.
193. Minoru Mori, Seiichi Uchida and Hitoshi Sakano, Dynamic Programming Matching with Global Features for Online Character Recognition, Proceedings of the 13th International Conference on Frontiers in Handwriting Recognition, 2012.09.
194. Yutaro Iwakiri, Soma Shiraishi, Yaokai Feng and Seiichi Uchida, On the Possibility of Instance-Based Stroke Recovery, Proceedings of the 13th International Conference on Frontiers in Handwriting Recognition, 2012.09.
195. Seiichi Uchida, Ryosuke Ishida, Akira Yoshida, Wenjie Cai and Yaokai Feng, Character Image Patterns as Big Data, Proceedings of the 13th International Conference on Frontiers in Handwriting Recognition, 2012.09.
196. Y. Furusawa, M. Imanishi, S. Hirata, S. Uchida, K. Nakano, K. Hayashi, Fluorescence Sensing Film for Odor Imaging, Proceedings of the 6th Asia-Pacific Conference on Transducers and Micro/Nano Technologies, 2012.07.
197. Masakazu Iwamura, Akira Horimatsu, Ryo Niwa, Koichi Kise, Seiichi Uchida, Shinichiro Omachi, Affine-invariant character recognition by progressive removing, Electrical Engineering in Japan (English translation of Denki Gakkai Ronbunshi), 10.1002/eej.22276, 180, 2, 55-63, 2012.07, Recognizing characters in scene images suffering from perspective distortion is a challenge. Although there are some methods to overcome this difficulty, they are time-consuming. In this paper, we propose a set of affine-invariant features and a new recognition scheme called "progressive removing" that can help reduce the processing time. Progressive removing gradually removes less feasible categories and skew angles by using multiple classifiers. We observed that progressive removing and the use of the affine invariant features reduced the processing time by about 60% in comparison to a trivial algorithm without decreasing the recognition rate. © 2012 Wiley Periodicals, Inc. Electr Eng Jpn, 180(2): 55-63, 2012; Published online in Wiley Online Library (). DOI 10.1002/eej.22276.
198. Soma Shiraishi, Yaokai Feng, Seiichi Uchida, A part-based skew estimation method, 10th IAPR International Workshop on Document Analysis Systems, DAS 2012 Proceedings - 10th IAPR International Workshop on Document Analysis Systems, DAS 2012, 10.1109/DAS.2012.7, 185-189, 2012.05, In this paper we propose a part-based skew estimation method which is more robust to larger varieties of text images, such as camera-captured scene images. Specifically, the skew angle at each local part of the input image is estimated independently by referring the local part of upright character images stored as a database. Then the global skew angle is estimated by aggregating the estimated local skews. The proposed method does not assume that characters are laid-out in straight lines and thus have more robustness to the varieties of text images than conventional methods. The experimental results show the advantage of the proposed method over the conventional methods under several conditions..
199. Minoru Mori, Seiichi Uchida, Hitoshi Sakano, How important is global structure for characters?, 10th IAPR International Workshop on Document Analysis Systems, DAS 2012 Proceedings - 10th IAPR International Workshop on Document Analysis Systems, DAS 2012, 10.1109/DAS.2012.41, 255-260, 2012.05, This paper studies the importance of the features that represent the global structure of character strokes to character recognition. Most existing character recognition methods based on character stroke features utilize a set or a sequence of local features such as xy-coordinates and local direction of strokes. This is natural from the viewpoint that each stroke is a trajectory and thus can be represented as a sequence of local features. This viewpoint, however, has a clear limitation in that local features cannot deal with global structure directly. For example, the sequence of local features cannot deal with the fact that the two end points of character "0" should be close to each other. In this paper we propose a simple and novel global feature that describes the global structure of the character shape of each class. We prove the importance of the global feature through a feature selection experiment. Specifically, we show that the global features are more often selected than local features to enhance classification accuracy under the AdaBoost-based machine learning framework. Recognition experiments using online numeral data show also that the use of global features improves recognition accuracy..
200. Michael Blumenstein, Umapada Pal, Seiichi Uchida, Message from general chair and program chairs, 10th IAPR International Workshop on Document Analysis Systems, DAS 2012 Proceedings - 10th IAPR International Workshop on Document Analysis Systems, DAS 2012, 10.1109/DAS.2012.55, xii-xiii, 2012.05.
201. Wang Song, Seiichi Uchida, Marcus Liwicki, Toward part-based document image decoding, 10th IAPR International Workshop on Document Analysis Systems, DAS 2012 Proceedings - 10th IAPR International Workshop on Document Analysis Systems, DAS 2012, 10.1109/DAS.2012.90, 266-270, 2012.05, Document image decoding (DID) is a trial to understand the contents of a whole document without any reference information about font, language, etc. Typically, DID approaches assume the correct segmentation of the document and some a priori knowledge about the language or the script. Unfortunately, this assumption will not hold if we deal with various documents, such as documents with various sized fonts, camera-captured documents, free-layout documents, or historical documents. In this paper, we propose a part-based character identification method where no segmentation into characters is necessary and no a priori information about the document is needed. The approach clusters similar key points and groups frequent neighboring key point clusters. Then a second iteration is performed, i.e., the groups are again clustered and optionally pairs frequent group clusters are detected. Our first experimental results on multi font-size documents look already very promising. We could find nearly perfect correspondences between characters and detected group clusters..
202. Asif Shahab, Faisal Shafait, Andreas Dengel and Seiichi Uchida, How Salient is Scene Text?, Proceedings of The 10th IAPR International Workshop on Document Analysis Systems (DAS2012, Gold Coast, Australia), 2012.03.
203. Wang Song, Marcus Liwicki, and Seiichi Uchida, Toward Part-based Document Image Decoding, Proceedings of The 10th IAPR International Workshop on Document Analysis Systems (DAS2012, Gold Coast, Australia), 2012.03.
204. Soma Shiraishi, Yaokai Feng and Seiichi Uchida, A Part-Based Skew Estimation Method, Proceedings of The 10th IAPR International Workshop on Document Analysis Systems (DAS2012, Gold Coast, Australia), 2012.03.
205. Minoru Mori, Seiichi Uchida and Hitoshi Sakano, How Important Is Global Structure for Characters?, Proceedings of The 10th IAPR International Workshop on Document Analysis Systems (DAS2012, Gold Coast, Australia), 2012.03.
206. Soma Shiraishi, Yaokai Feng and Seiichi Uchida, Part-Based Skew Estimation for Mathematical Expressions, Proceedings of The International Workshop on "Digitization and E-Inclusion in Mathematics and Science 2012 (DEIMS12, Tokyo, Japan), 2012.02.
207. Hirotaka Matsuo, Yudai Furusawa, Masashi Imanishi, Seiichi Uchida, Kenshi Hayashi, Optical odor imaging by fluorescence probes, Journal of Robotics and Mechatronics, 24, 1, 47-54, 2012.02, Odor gas detection is important for the detection of explosives, environmental sensing, biometrics, foodstuffs and a comfortable life. Such odor-source localizations is an active research area for robotics. In this study, we tried to detect odor chemicals with an optical method that can be applied for the spatiotemporal detection of odor. We used four types of fluorescence dyes; tryptophan, quinine sulfate, acridine orange, and 1-anilinonaphthalene-8-sulfonate (ANS). As analyses, we measured the following four odor chemicals, 2-furaldehyde, vanillin, acetophenone, and benzaldehyde. The fluorescence-quenching mechanism of PET (Photoinduced Electron Transfer) or FRET (Fluorescence Resonance Electron Transfer), which occur between fluorescence dyes and odor compounds, could prevent unintended detection of various odorants that is caused by their unspecific adsorption onto the detecting materials. The fluorescence changes were then observed. Thus, we could detect the odor substances through fluorescent quenching by using the fluorescence dyes. Odor information could be obtained by response patterns across all the fluorescence dyes. Moreover, we captured odor images with a cooled CCD camera. Shapes of the targets that emitted odor could be roughly recognized by the odor-shape images. From the spatiotemporal images of odors, twodimensional odor expanse could be obtained..
208. Hirotaka Matsuo, Yudai Furusawa, Masashi Imanishi, Seiichi Uchida, and Kenshi Hayashi, Optical Odor Imaging by Fluorescence Probes, Journal of Robotics and Mechatoronics, 2012.01.
209. Seiichi Uchida, Satoshi Hokahori, Yaokai Feng, Analytical dynamic programming matching, 12th European Conference on Computer Vision, ECCV 2012 Computer Vision, ECCV 2012 - Workshops and Demonstrations, Proceedings, 10.1007/978-3-642-33863-2_10, 92-101, 2012.01, In this paper, we show that the truly two-dimensional elastic image matching problem can be solved analytically using dynamic programming (DP) in polynomial time if the problem is formulated as a maximum a posteriori problem using Gaussian distributions for the likelihood and prior. After giving the derivation of the analytical DP matching algorithm, we evaluate its performance on handwritten character images containing various nonlinear deformations, and compare other elastic image matching methods..
210. Asif Shahab, Faisal Shafait, Andreas Dengel, Seiichi Uchida, How salient is scene text?, 10th IAPR International Workshop on Document Analysis Systems, DAS 2012 Proceedings - 10th IAPR International Workshop on Document Analysis Systems, DAS 2012, 10.1109/DAS.2012.42, 317-321, 2012, Computational models of visual attention use image features to identify salient locations in an image that are likely to attract human attention. Attention models have been quite effectively used for various object detection tasks. However, their use for scene text detection is under-investigated. As a general observation, scene text often conveys important information and is usually prominent or salient in the scene itself. In this paper, we evaluate four state-of-the-art attention models for their response to scene text. Initial results indicate that saliency maps produced by these attention models can be used for aiding scene text detection algorithms by suppressing non-text regions..
211. Rong Huang, Shinpei Oba, Shivakumara Palaiahnakote, Seiichi Uchida, Scene character detection and recognition based on multiple hypotheses framework, 21st International Conference on Pattern Recognition, ICPR 2012 ICPR 2012 - 21st International Conference on Pattern Recognition, 717-720, 2012, To handle the diversity of scene characters, we propose a multiple hypotheses framework which consists of an image operator set module, an optical character recognition (OCR) module, and an integration module. Image operators detect multiple suspicious character areas. The OCR engine is then applied to each detected area and returns multiple candidates with weight values for future integration. Without the aid of heuristic constraints on area, aspect ratio or color etc., the integration module prunes the redundant detection and pads the missing detection based on the outputs of OCR. The experimental results demonstrate that the whole multiple hypotheses outperforms each operator's hypotheses and be comparable with existing methods in terms of recall, precision, F-measure and recognition rate..
212. Seiichi Uchida, Toru Sasaki, Feng Yaokai, A generative model for handwritings based on enhanced feature desynchronization, 11th International Conference on Document Analysis and Recognition, ICDAR 2011 Proceedings - 11th International Conference on Document Analysis and Recognition, ICDAR 2011, 10.1109/ICDAR.2011.124, 589-593, 2011.12, A new generative model of handwriting patterns is proposed for interpreting their deformations. The model is based on feature desynchronization, which is a coupling process of x and y coordinate features of different timings. By changing the timings to be coupled, the model can generate various deformed patterns from a single pattern. The model is further enhanced by incorporating an adaptive rotation at each timing for increasing the variety of deformed patterns. An important fact is that this enhanced desynchronization model can be interpreted intuitively as a deformation process in actual handwriting. Experimental results showed that the model can generate various handwriting patterns close to actual deformed patterns..
213. Seiichi Uchida, Yuki Shigeyoshi, Yasuhiro Kunishige, Yaokai Feng, A keypoint-based approach toward scenery character detection, 11th International Conference on Document Analysis and Recognition, ICDAR 2011 Proceedings - 11th International Conference on Document Analysis and Recognition, ICDAR 2011, 10.1109/ICDAR.2011.168, 819-823, 2011.12, This paper proposes a new approach toward scenery character detection. This is a key point-based approach where local features and a saliency map are fully utilized. Local features, such as SIFT and SURF, have been commonly used for computer vision and object pattern recognition problems, however, they have been rarely employed in character recognition and detection problems. Local feature, however, is similar to directional features, which have been employed in character recognition applications. In addition, local feature can detect corners and thus it is suitable for detecting characters, which are generally comprised of many corners. For evaluating the performance of the local feature, an experimental result was done and its results showed that SURF, i.e., a simple gradient feature, can detect about 70% of characters in scenery images. Then the saliency map was employed as an additional feature to the local feature. This trial is based on the expectation that scenery characters are generally printed to be salient and thus higher salient area will have a higher probability to be a character area. An experimental result showed that this expectation was reasonable and we can have better discrimination accuracy with the saliency map..
214. Wang Song, Seiichi Uchida, Marcus Liwicki, Comparative study of part-based handwritten character recognition methods, 11th International Conference on Document Analysis and Recognition, ICDAR 2011 Proceedings - 11th International Conference on Document Analysis and Recognition, ICDAR 2011, 10.1109/ICDAR.2011.167, 814-818, 2011.12, The purpose of this paper is to introduce three part-based methods for handwritten character recognition and then compare their performances experimentally. All of those methods decompose handwritten characters into "parts". Then some recognition processes are done in a part-wise manner and, finally, the recognition results at all the parts are combined via voting to have the recognition result of the entire character. Since part-based methods do not rely on the global structure of the character, we can expect their robustness against various deformations. Three voting methods have been investigated for the combination: single voting, multiple voting, and class distance. All of them use different strategies for voting. Experimental results on the MNIST database showed the relative superiority of the class distance method and the robustness of the multiple voting method against the reduction of training set..
215. Wang Song, Seiichi Uchida, Marcus Liwicki, Look inside the world of parts of handwritten characters, 11th International Conference on Document Analysis and Recognition, ICDAR 2011 Proceedings - 11th International Conference on Document Analysis and Recognition, ICDAR 2011, 10.1109/ICDAR.2011.161, 784-788, 2011.12, Part-based recognition is expected to be robust in difficult handwritten character recognition tasks. This is because part-based recognition is based on aggregation of independent recognition results at individual local parts without considering their global relations and thus is robust against various deformations, such as partial occlusion, overlap, broken stroke, etc. Since part-based recognition is a new approach, there are still several open problems toward its practical use. For example, compared with entire images, local parts are more ambiguous, i.e., less discriminative. For better recognition accuracy and less computations, we need to know the characteristics of local parts and then, for example, discard less discriminative parts. The purpose of this paper is to conduct some experiments in order to observe and analyze how the local parts of multiple classes are distributed in feature spaces. By handling parts appropriately based on the analysis, we will be able to enhance the usefulness of the part-based method..
216. Marcus Liwicki, Yoshida Akira, Seiichi Uchida, Masakazu Iwamura, Shinichiro Omachi, Koichi Kise, Reliable online stroke recovery from offline data with the data-embedding pen, 11th International Conference on Document Analysis and Recognition, ICDAR 2011 Proceedings - 11th International Conference on Document Analysis and Recognition, ICDAR 2011, 10.1109/ICDAR.2011.278, 1384-1388, 2011.12, In this paper we propose a complete system for online stroke recovery from offline data. The key idea of our approach is to use a novel pen device which is able to embed meta information into the ink during writing the strokes. This pen-device overcomes the need to get access to any memory on the pen when trying to recover the information, which is especially useful in multi-writer or multi-pen scenarios. The actual data-embedding is achieved by an additional ink dot sequence along a handwritten pattern during writing. We design the ink-dot sequence in such a way that it is possible to retrieve the writing direction from a scanned image. Furthermore, we propose novel processing steps in order to retrieve the original writing direction and finally the embedded data. In our experiments we show that we can reliably recover the writing direction of various patterns. Our system is able to determine the writing direction of straight lines, simple patterns with crossings (e.g., "x" and "II"), and even more complex patterns like handwritten words and symbols..
217. Yasuhiro Kunishige, Yaokai Feng, Seiichi Uchida, Scenery character detection with environmental context, 11th International Conference on Document Analysis and Recognition, ICDAR 2011 Proceedings - 11th International Conference on Document Analysis and Recognition, ICDAR 2011, 10.1109/ICDAR.2011.212, 1049-1053, 2011.12, For scenery character detection, we introduce environmental context, which is modeled by scene components, such as sky and building. Environmental context is expected to regulate the probability of character existence at a specific region in a scenery image. For example, if a region looks like a part of a building, the region has a higher probability than another region like a part of the sky. In this paper, environmental context is represented by state-of-the-art texture and color features and utilized in two different ways. Through experimental results, it was clearly shown that the environmental context has an effect of improving detection accuracy..
218. Seiichi Uchida, Wenjie Cai, Akira Yoshida, Yaokai Feng, Watching pattern distribution via massive character recognition, 21st IEEE International Workshop on Machine Learning for Signal Processing, MLSP 2011 2011 IEEE International Workshop on Machine Learning for Signal Processing - Proceedings of MLSP 2011, 10.1109/MLSP.2011.6064640, 2011.12, The purpose of this paper is to analyze how image patterns distribute inside their feature space. For this purpose, 832,612 manually ground-truthed handwritten digit patterns are used. Use of character patterns instead of general visual object patterns is very essential for our purpose. First, since there are only 10 classes for digits, it is possible to have an enough number of patterns per class. Second, since the feature space of small binary character images is rather compact, it is easier to observe the precise pattern distribution with a fixed number of patterns. Third, the classes of character patterns can be defined far more clearly than visual objects. Through nearest neighbor analysis on 832, 612 patterns, their distribution in the 32 x 32 binary feature space is observed quantitatively and qualitatively. For example, the visual similarity of nearest neighbors and the existence of outliers, which are surrounded by patterns from different classes, are observed..
219. Seiichi Uchida, Wenjie Cai, Akira Yoshida, Yaokai Feng, Watching Pattern Distribution via Massive Character Recognition, 2011 IEEE International Workshop on Machine Learning for Signal Processing (MLSP2011, Beijing, China), 2011.09.
220. Seiichi Uchida, Toru Sasaki and Yaokai Feng, A Generative Model for Handwritings Based on Enhanced Feature Desynchronization, Proceedings of The 11th International Conference on Document Analysis and Recognition (ICDAR 2011, Beijing, China), 2011.09.
221. Yasuhiro Kunishige, Yaokai Feng and Seiichi Uchida, Scenery Character Detection with Environmental Context, Proceedings of The 11th International Conference on Document Analysis and Recognition (ICDAR 2011, Beijing, China), 2011.09.
222. Song Wang, Seiichi Uchida and Marcus Liwicki, Look Inside the World of Parts of Handwritten Characters, Proceedings of The 11th International Conference on Document Analysis and Recognition (ICDAR 2011, Beijing, China), 2011.09.
223. Song Wang, Seiichi Uchida and Marcus Liwicki, Comparative Study of Part-Based Handwritten Character Recognition Methods, Proceedings of The 11th International Conference on Document Analysis and Recognition (ICDAR 2011, Beijing, China), 2011.09.
224. Seiichi Uchida, Yuki Shigeyoshi, Yasuhiro Kunishige and Yaokai Feng, A Keypoint-Based Approach Toward Scenery Character Detection, Proceedings of The 11th International Conference on Document Analysis and Recognition (ICDAR 2011, Beijing, China), 2011.09.
225. Marcus Liwicki, Yoshida Akira, Seiichi Uchida, Masakazu Iwamura, Shinichiro Omachi and Koichi Kise, Reliable Online Stroke Recovery from Offline Data with the Data-Embedding Pen, Proceedings of The 11th International Conference on Document Analysis and Recognition (ICDAR 2011, Beijing, China), 2011.09.
226. Soma Shiraishi, Yaokai Feng, Seiichi Uchida, A new approach for instance-based skew estimation, 15th International Conference on Knowledge-Based and Intelligent Information and Engineering Systems, KES 2011 Knowledge-Based and Intelligent Information and Engineering Systems - 15th International Conference, KES 2011, Proceedings, 10.1007/978-3-642-23866-6_21, 195-203, 2011.09, This paper proposes a new approach to a method to estimate a skew angle of a rotated document image. This is realized by using Speeded-Up Robust Features (SURF), and the goal is that it enables the image to be rotated back to the correct orientation. SURF detects a number of keypoints both from the reference image on which a set of standard alphabets (e.g. letter eaf through ezf in a certain font) are written, and the image of the rotated document. Two nearest features each from the reference image and the input image are compared to decide to how many degrees the feature in the input image is rotated. Finally the skew angle of the whole input image( the global skew angle) is decided by the majority of the total votes of angles that have been calculated as mentioned above..
227. Masakazu Iwamura, Akira Horimatsu, Ryo Niwa, Koichi Kise, Seiichi Uchida, Shinichiro Omachi, Affine invariant character recognition by progressive removing, IEEJ Transactions on Industry Applications, 10.1541/ieejias.131.873, 131, 7, 873-879, 2011.09, Recognizing characters in scene images suffering from perspective distortion is a challenge. Although there are some methods to overcome this difficulty, they are time-consuming. In this paper, we propose a set of affine invariant features and a new recognition scheme called "progressive removing" that can help reduce the processing time. Progressive removing gradually removes less feasible categories and skew angles by using multiple classifiers. We observed that progressive removing and the use of the affine invariant features reduced the processing time by about 60% in comparison to a trivial one without decreasing the recognition rate..
228. Akira Yoshida, Marcus Liwichi, Seiichi Uchida, Masakazu Iwamura, Shinichiro Omachi, Koichi Kise, Handwriting on paper as a cybermedium, 15th International Conference on Knowledge-Based and Intelligent Information and Engineering Systems, KES 2011 Knowledge-Based and Intelligent Information and Engineering Systems - 15th International Conference, KES 2011, Proceedings, 10.1007/978-3-642-23866-6_22, 204-211, 2011.09, In this paper, we report recent work of the data-embedding pen, which adds an ink-dot sequence along a handwritten pattern during writing. The ink-dot sequence represents some information, such as writer's name, date of writing, and URL. This information drastically increases the value of handwriting on a paper. The embedded information can be extracted from the handwritten pattern by image processing techniques and a stroke recovery technique. Consequently, we can augment the handwritten pattern by the data-embedding pen to carry arbitrary information..
229. Kou Yamada, Seiichi Uchida, Rin-Ichiro Taniguchi, Object tracking with RFID, IEEJ Transactions on Industry Applications, 10.1541/ieejias.131.441, 131, 4, 2011.09, This paper reports a new method for visual tracking of humans using active RFID technology. Previous studies were based on the assumption that the radio intensity from an RFID tag will be linearly proportional to the distance between the tag and the antenna or will remain unchanged; however, in reality, the intensity fluctuates significantly and changes drastically with a small change in the environment. The proposed method helps to overcome this problem by using only accurate binary information that reveals whether the target person is close to the antenna. Several experimental results have shown that the information from the RFID tag was useful for reliable tracking of humans..
230. Wenjie Cai, Yaokai Feng, Seiichi Uchida, Massive character recognition with a large ground-truthed database, 26th Annual ACM Symposium on Applied Computing, SAC 2011 26th Annual ACM Symposium on Applied Computing, SAC 2011, 10.1145/1982185.1982241, 240-244, 2011.06, In character recognition, multiple prototype classifiers, where multiple patterns are prepared as representative patterns of each class, have often been employed to improve recognition accuracy. Our question is how we can improve the recognition accuracy by increasing prototypes massively in the multiple prototype classifier. In this paper, we will answer this question through several experimental analyses, using a simple 1-nearest neighbor (1-NN) classifier and about 550,000 manually labeled handwritten numeral patterns. The analysis results under the leave-one-out evaluation showed not only a simple fact that more prototypes provide fewer recognition errors, but also a more important fact that the error rate decreases approximately to 40% by increasing the prototypes 10 times. The analysis results also showed other phenomena in massive character recognition, such that the NN prototypes become visually closer to the input pattern by increasing the prototypes..
231. A. Nedzved, O. Nedzved, Sergey Ablameyko, Seiichi Uchida, Object Extraction at Nano-Surface Images, Proceedings of The Eleventh International Conference on Pattern Recognition and Information Processing (PRIP2011, Minsk, Belarus), 2011.05.
232. Wenjie Cai, Yaokai Feng and Seiichi Uchida, Massive Character Recognition with a Large Ground-Truthed Database, Proceedings of 26th Symposium On Applied Computing, 2011.03.
233. Seiichi Uchida, Ikko Fujimura, Hiroki Kawano, Yaokai Feng, Analytical dynamic programming tracker, 10th Asian Conference on Computer Vision, ACCV 2010 Computer Vision, ACCV 2010 - 10th Asian Conference on Computer Vision, Revised Selected Papers, 10.1007/978-3-642-19315-6_23, 296-309, 2011.03, Visual tracking is formulated as an optimization problem of the position of a target object on video frames. This paper proposes a new tracking method based on dynamic programming (DP). Conventional DP-based tracking methods have utilized DP as an efficient breadth-first search algorithm. Thus, their computational complexity becomes prohibitive if the search breadth becomes large according to the increase of the number of parameters to be optimized. In contrast, the proposed method can avoid this problem by utilizing DP as an analytical solver rather than the conventional breadth-first search algorithm. In addition to experimental evaluations, it will be revealed that the proposed method has a close relation to the well-known KLT tracker..
234. Wenjie Cai, Seiichi Uchida, Hiroaki Sakoe, Toward forensics by stroke order variation - Performance evaluation of stroke correspondence methods, 4th International Workshop on Computational Forensics, IWCF 2010 Computational Forensics - 4th International Workshop, IWCF 2010, Revised Selected Papers, 10.1007/978-3-642-19376-7_4, 43-55, 2011.03, We consider personal identification using stroke order variations of online handwritten character patterns, which are written on, e.g., electric tablets. To extract the stroke order variation of an input character pattern, it is necessary to establish the accurate stroke correspondence between the input pattern and the reference pattern of the same category. In this paper we compare five stroke correspondence methods: the individual correspondence decision (ICD), the cube search (CS), the bipartite weighted matching (BWM), the stable marriage (SM), and the deviation-expansion model (DE). After their brief review, they are experimentally compared quantitatively by not only their stroke correspondence accuracy but also character recognition accuracy. The experimental results showed the superiority CS and BWM over ICD, SM and DE..
235. Akihiro Mori, Seiichi Uchida, Ryo Kurazume, Rin-Ichiro Taniguchi, Tsutomu Hasegawa, Automatic construction of gesture network for gesture recognition, 2010 IEEE Region 10 Conference, TENCON 2010 TENCON 2010 - 2010 IEEE Region 10 Conference, 10.1109/TENCON.2010.5686549, 923-928, 2010.12, This paper is concerned with automatic construction algorithm for gesture network. Gesture network is a network model of gestures for gesture recognition, especially early recognition and motion prediction. Manual construction of gesture network is inefficient, and thus its automatic construction method is expected; this is because gesture network has to be constructed, whenever target gestures are changed. This paper proposes an automatic construction algorithm for gesture network by logical DP matching. The experiment was conducted for evaluating the performance of the gesture network constructed automatically. The experimental result indicated that the proposed automatic construction algorithm for gesture network can be alternative of manual construction..
236. Marcus Liwicki, Seiichi Uchida, Masakazu Iwamura, Shinichiro Omachi, Koichi Kise, Embedding Meta-information in handwriting - Reed-solomon for reliable error correction, 12th International Conference on Frontiers in Handwriting Recognition, ICFHR 2010 Proceedings - 12th International Conference on Frontiers in Handwriting Recognition, ICFHR 2010, 10.1109/ICFHR.2010.127, 51-56, 2010.12, In this paper a more compact and more reliable coding scheme for the data-embedding pen is proposed. The data-embedding pen produces an additional ink-dot sequence along a handwritten pattern during writing. The ink-dot sequence represents, for example, meta-information (such as the writer's name and the date of writing) and thus drastically increases the value of the handwriting on a physical paper. There is no need to get access to any memory on the pen to recover the information, which is especially useful in multi-writer or multi-pen scenarios. In this paper we focus on the compactness of the encoded information. The aim of this paper is to encode as much information as possible in short stroke sequences. In our experiments we show that we can embed more information in shorter strokes than in previous work. In straight lines as short as 5 cm, 32 bits can successfully be embedded. Furthermore, the new encoding scheme also works reliably on more complex patterns..
237. Seiichi Uchida, Marcus Liwicki, Part-based recognition of handwritten characters, 12th International Conference on Frontiers in Handwriting Recognition, ICFHR 2010 Proceedings - 12th International Conference on Frontiers in Handwriting Recognition, ICFHR 2010, 10.1109/ICFHR.2010.90, 545-550, 2010.12, In the part-based recognition method proposed in this paper, a handwritten character image is represented by just a set of local parts. Then, each local part of the input pattern is recognized by a nearest-neighbor classifier. Finally, the category of the input pattern is determined by aggregating the local recognition results. This approach is opposed to conventional character recognition approaches which try to benefit from the global structure information as much as possible. Despite a pessimistic expectation, we have reached recognition rates much higher than 90% for a digit recognition task. In this paper we provide a detailed analysis in order to understand the results and find the merits of the local approach..
238. Kazumasa Iwata, Koichi Kise, Masakazu Iwamura, Seiichi Uchida, Shinichiro Omachi, Tracking and retrieval of pen tip positions for an intelligent camera pen, 12th International Conference on Frontiers in Handwriting Recognition, ICFHR 2010 Proceedings - 12th International Conference on Frontiers in Handwriting Recognition, ICFHR 2010, 10.1109/ICFHR.2010.50, 277-282, 2010.12, This paper presents a method of recovering digital ink for an intelligent camera pen, which is characterized by the functions that (1) it works on ordinary paper and (2) if an electronic document is printed on the paper the recovered digital ink is associated with the document. Two technologies called paper fingerprint and document image retrieval are integrated for realizing the above functions. The key of the integration is the introduction of image mosaicing and fast retrieval of previously seen fingerprints based on hashing of SURF local features. From the experimental results of 50 handwritings, we have confirmed that the proposed method is effective to recover and locate the digital ink from the handwriting on a physical paper..
239. Seiichi Uchida, Ikko Fujimura, Hiroki Kawano, Yaokai Feng, Analytical Dynamic Programming Tracker, Proceedings of 10th Asian Conference on Computer Vision, 2010.11.
240. Akihiro Mori, Seiichi Uchida, Ryo Kurazume, Rin-ichiro Taniguchi, Tsutomu Hasegawa, Automatic Construction of Gesture Network for Gesture Recognition, Proceedings of IEEE TENCON2010, 2010.11.
241. Wenjie Cai, Seiichi Uchida, Hiroaki Sakoe, Toward Forensics by Stroke-Order Variation --- Performance Evaluation of Stroke Correspondence Methods, Proceedings of 4th International Workshop on Computational Forensics, 2010.11.
242. Seiichi Uchida and Marcus Liwicki, Part-Based Recognition of Handwritten Characters, Proceedings of The 12th International Conference on Frontiers in Handwriting Recognition, 2010.11.
243. Marcus Liwicki, Seiichi Uchida, Masakazu Iwamura, Shinichiro Omachi and Koichi Kise, Embedding Meta-Information in Handwriting — Reed-Solomon for Reliable Error Correction, Proceedings of The 12th International Conference on Frontiers in Handwriting Recognition, 2010.11.
244. Kazumasa Iwata, Koichi Kise, Masakazu Iwamura, Seiichi Uchida and Shinichiro Omachi, Tracking and Retrieval of Pen Tip Positions for an Intelligent Camera Pen, Proceedings of The 12th International Conference on Frontiers in Handwriting Recognition, 2010.11, This paper presents a method of recovering digital ink for an intelligent camera pen, which is characterized by the functions that (1) it works on ordinary paper and (2) if an electronic document is printed on the paper the recovered digital ink is associated with the document. Two technologies called paper fingerprint and document image retrieval are integrated for realizing the above functions. The key of the integration is the introduction of image mosaicing and fast retrieval of previously seen fingerprints based on hashing of SURF local features. From the experimental results of 50 handwritings, we have confirmed that the proposed method is effective to recover and locate the digital ink from the handwriting on a physical paper..
245. Seiichi Uchida, Marcus Liwicki, Analysis of local features for handwritten character recognition, 2010 20th International Conference on Pattern Recognition, ICPR 2010 Proceedings - 2010 20th International Conference on Pattern Recognition, ICPR 2010, 10.1109/ICPR.2010.479, 1945-1948, 2010.11, This paper investigates a part-based recognition method of handwritten digits. In the proposed method, the global structure of digit patterns is discarded by representing each pattern by just a set of local feature vectors. The method is then comprised of two steps. First, each of J local feature vectors of a target pattern is recognized into one of ten categories ("0"-"9") by the nearest neighbor discrimination with a large database of reference vectors. Second, the category of the target pattern is determined by the majority voting on the J local recognition results. Despite a pessimistic expectation, we have reached recognition rates much higher than 90% for the task of digit recognition..
246. Toru Wakahara, Seiichi Uchida, Hierarchical decomposition of handwriting deformation vector field for improving recognition accuracy, 2010 20th International Conference on Pattern Recognition, ICPR 2010 Proceedings - 2010 20th International Conference on Pattern Recognition, ICPR 2010, 10.1109/ICPR.2010.459, 1860-1863, 2010.11, This paper addresses the problem of how to extract, describe, and evaluate handwriting deformation from the deterministic viewpoint for improving recognition accuracy. The key ideas are threefold. The first is to extract handwriting deformation vector field (DVF) between a pair of input and target images by 2D warping. The second is to hierarchically decompose the DVF by a parametric deformation model of global/local affine transformation, where local affine transformation is iteratively applied to the DVF by decreasing window sizes. The third is to accept only low-order deformation components as natural, within-class handwriting deformation. Experiments using the handwritten numeral database IPTP CDROM1B show that correlation-based matching absorbing components of global affine transformation and local affine transformation up to the 3rd order achieved a higher recognition rate of 92.1% than that of 87.0% obtained by original 2D warping..
247. Toru Wakahara and Seiichi Uchida, Hierarchical Decomposition of Handwriting Deformation Vector Field for Improving Recognition Accuracy, Proceedings of The 20th International Conference on Pattern Recognition, 2010.08.
248. Seiichi Uchida and Marcus Liwicki, Analysis of Local Features for Handwritten Character Recognition, Proceedings of The 20th International Conference on Pattern Recognition, vol.129, no.5, 2010.08.
249. Marcus Liwicki, Seiichi Uchida, Masakazu Iwamura, Shinichiro Omachi, Koichi Kise, Data-embedding pen-augmenting ink strokes with meta-information, 2010 IAPR Workshop on Document Analysis Systems, DAS 2010 Proceedings of the 9th IAPR International Workshop on Document Analysis Systems, DAS '10, 10.1145/1815330.1815336, 43-51, 2010.08, In this paper we present the first operational version of the data-embedding pen. During writing a pattern, this pen produces an additional ink-dot sequence along the ink stroke of the pattern. The ink-dot sequence represents, for exam-ple, meta-information (such as the writer's name and the date of writing) and thus drastically increases the value of the handwriting on a physical paper. Since the information is placed on the paper, it can be extracted just by scanning or photographing the paper. There is no need to get access to any memory on the pen to recover the information. This is useful especially in multi-writer or multi-pen scenarios. The experiments using an encoding scheme and a decoding algorithm showed very promising results. For example, it was proved that we can embed 28 or more bits of informa-tion on simple handwritten patterns and decode them with a high reliability..
250. Koichi Kise, Masakazu Iwamura, Megumi Chikano, Seiichi Uchida, Kazumasa Iwata, Shinichiro Omachi, Expansion of queries and databases for improving the retrieval accuracy of document portions
An application to a camera-pen system, 2010 IAPR Workshop on Document Analysis Systems, DAS 2010 Proceedings of the 9th IAPR International Workshop on Document Analysis Systems, DAS '10, 10.1145/1815330.1815370, 309-316, 2010.08, This paper presents a method of improving the accuracy of document image retrieval focusing on the application to a camera-pen system. In a camera-pen system, document image retrieval is employed for locating the pen-tip position on a page. A serious problem is that since the camera is mounted close to the pen-tip, the camera captures only a tiny portion of the page and the resultant image is under severe perspective distortion, resulting in lowering the retrieval accuracy. To solve this problem, we propose new geometrically invariant features as well as expansion techniques which increase the number of index features of either the database or the query images. Prom the experimental results, it has been found that the query expansion technique with features by combining affine and perspective invariants allows us the best performance that improves the accuracy of a baseline method more than 27%..
251. Yaokai Feng, Seiichi Uchida, How to design Kansei retrieval systems?, 11th International Conference on Web-Age Information Management, WAIM 2010 Web-Age Information Management - 11th International Conference, WAIM 2010, Proceedings, 10.1007/978-3-642-14246-8_40, 405-416, 2010.08, All of the information retrieval (IR) systems try to retrieve the information that users really want. For this purpose, users have to exactly express and submit their requirements to the systems. However, how to reflect user's subjectivities is a hard problem. When a user wants to search for something, for instance a passenger car or a costume, he/she normally has a kind of feeling such as "graceful and looks intelligent, but not so expensive." Many researchers call this feeling as "Kansei" in many researches, which means the user's psychological feeling as well as the physiological issues. Let's see another example. Image retrieval systems having an ability to handle subjective expressions are useful especially when the users, who have no knowledge about contents of the image database, try to use some Kansei words (e.g., " beautiful", "calm", etc.) to retrieve unknown images. Unfortunately, traditional information retrieval systems cannot efficiently deal with the user-given search requests with Kansei words. Thus, in many application systems, how to deal with Kansei words from different users has become an important issue that the system designers have to confront. In this background, "Kansei engineering" has become one of the hot topics in IR field. In addition, many Kansei retrieval systems have been implemented. However, all of the existing Kansei retrieval systems aim at specific applications. Thus, they are very different from each other. In this paper, based on the many investigations and analysis on the existing systems, we propose a general method for designing efficient Kansei retrieval systems, which has not been done in this field. Our proposal mainly includes a general flow for designing Kansei retrieval systems and a discussion on how to speed up the Kansei retrieval processes using multidimensional indexes. We hope that this paper will be able to help the designers who are planning to design Kansei retrieval systems..
252. Koichi Kise, Megumi Chikano, Kazumasa Iwata, Masakazu Iwamura, Seiichi Uchida and Shinichiro Omachi, Expansion of Queries and Databases for Improving the Retrieval Accuracy of Document Portions, Proceedings of The Ninth International Workshop on Document Analysis Systems, 2010.06.
253. Marcus Liwicki, Seiichi Uchida, Masakazu Iwamura, Shinichiro Omachi and Koichi Kise, Data-Embedding Pen - Augmenting Ink Strokes with Meta-Information, Proceedings of The Ninth International Workshop on Document Analysis Systems, 2010.06.
254. Akio Fujiyoshi, Masakazu Suzuki, Seiichi Uchida, Grammatical verification for mathematical formula recognition based on context-free tree grammar, Mathematics in Computer Science, 10.1007/s11786-010-0023-8, 3, 3, 279-298, 2010.05, This paper proposes the use of a formal grammar for the verification of mathematical formulae for a practical mathematical OCR system. Like a C compiler detecting syntax errors in a source file, we want to have a verification mechanism to find errors in the output of mathematical OCR. A linear monadic context-free tree grammar (LM-CFTG) is employed as a formal framework to define "well-formed" mathematical formulae. A cubic time parsing algorithm for LM-CFTGs is presented. For the purpose of practical evaluation, a verification system for mathematical OCR is developed, and the effectiveness of the system is demonstrated by using the ground-truthed mathematical document database InftyCDB-1 and a misrecognition database newly constructed for this study..
255. Akio Fujiyoshi, Masakazu Suzuki and Seiichi Uchida, Grammatical Verification for Mathematical Formula Recognition Based on Context-Free Tree Grammar, Mathematics in Computer Science, vol.10, no.4, pp.559-567, 2010.03.
256. Seiichi Uchida, Marcus Liwicki, Masakazu Iwamura, Koichi Kise, Shinichiro Omachi, Digital pen, Kyokai Joho Imeji Zasshi/Journal of the Institute of Image Information and Television Engineers, 10.3169/itej.64.293, 64, 3, 293-298, 2010.03.
257. Real-Time Nonlinear FEM-Based Simulator for Deforming Volume Model of Soft Organ by Neural Network.
258. Walaa Aly, Seiichi Uchida and Masakazu Suzuki, Extract Baseline Information Using Support Vector Machine, Proceedings of The 9th Asian Symposium on Computer Mathematics, 2009.12.
259. Trachea and Esophagus Classification by AdaBoost.
260. Kazumasa Iwata, Koichi Kise, Tomohiro Nakai, Masakazu Iwamura, Seiichi Uchida, Shinichiro Omachi, Capturing digital ink as retrieving fragments of document images, ICDAR2009 - 10th International Conference on Document Analysis and Recognition ICDAR2009 - 10th International Conference on Document Analysis and Recognition, 10.1109/ICDAR.2009.192, 1236-1240, 2009.12, This paper presents a new method of capturing digital ink for pen-based computing. Current technologies such as tablets, ultrasonic and the Anoto pens rely on special mechanisms for locating the pen tip, which result in limiting the applicability. Our proposal is to ease this problem - a camera pen that allows us to write on ordinary paper for capturing digital ink. A document image retrieval method called LLAH is tuned to locate the pen tip efficiently and accurately on the coordinates of a document only by capturing its tiny fragment. In this papeic we report some results on captured digital ink as well as to evaluate their quality..
261. Seiichi Uchida, Ryoji Hattori, Masakazu Iwamura, Shinichiro Omachi, Koichi Kise, Conspicuous character patterns, ICDAR2009 - 10th International Conference on Document Analysis and Recognition ICDAR2009 - 10th International Conference on Document Analysis and Recognition, 10.1109/ICDAR.2009.196, 16-20, 2009.12, Detection of characters in scenery images is often a very difficult problem. Although many researchers have tackled this difficult problem and achieved a good performance, it is still difficult to suppress many false alarms andalthough missings. This paper investigates a conspicuous character pattern, which is a special pattern designed for easier detection. In order to have an example of the conspicuous character pattern, we select a character font with a larger distance from a non-character pattern distribution and, simultaneously, with a smaller distance from a character pattern distribution. Experimental results showed that the character font selected by this method is actually more conspicuous (i.e., detected more easily) than other fonts..
262. Toru Wakahara, Seiichi Uchida, Hierarchical decomposition of handwriting deformation vector field using 2D warping and global/local affine transformation, ICDAR2009 - 10th International Conference on Document Analysis and Recognition ICDAR2009 - 10th International Conference on Document Analysis and Recognition, 10.1109/ICDAR.2009.33, 1141-1145, 2009.12, This paper addresses the basic problem of how to extract, describe, and evaluate handwriting deformation from not the statistical but the deterministic viewpoint. The key ideas are threefold. The first idea is to apply 2D warping to extraction of handwriting deformation vector field (DVF) between a pair of input and target images. The second idea is to hierarchically decompose the DVF by a parametric deformation model of global/local affine transformation. As a result, the DVF is expressed by a series of deformation components each of which is characterized by a window size of local affine transformation. The third idea is interrupting of the series of deformation components to obtain natural, reasonable handwriting deformation. Experiments using the handwritten numeral database IPTP CDROM1B show that 31.1% of the handwriting DVF is expressed by global affine transformation, and the subsequent few local affine transformations successfully discriminate natural handwriting deformation from unnatural one..
263. Walaa Aly, Seiichi Uchida, Akio Fujiyoshi, Masakazu Suzuki, Statistical classification of spatial relationships among mathematical symbols, ICDAR2009 - 10th International Conference on Document Analysis and Recognition ICDAR2009 - 10th International Conference on Document Analysis and Recognition, 10.1109/ICDAR.2009.90, 1350-1354, 2009.12, In this paper, a statistical decision method for automatic classification of spatial relationships between each adjacent pair is proposed. Each pair is composed of mathematical symbols and/or alphabetical characters. Special treatment of mathematical symbols with variable size is important. This classification is important to recognize an accurate structure analysis module of math OCR. Experimental results on a very large database showed that the proposed method worked well with an accuracy of 99.57% by two important geometric feature relative size and relative position..
264. Akio Fujiyoshi, Masakazu Suzuki, Seiichi Uchida, Syntactic detection and correction of misrecognitions in mathematical OCR, ICDAR2009 - 10th International Conference on Document Analysis and Recognition ICDAR2009 - 10th International Conference on Document Analysis and Recognition, 10.1109/ICDAR.2009.150, 1360-1364, 2009.12, This paper proposes a syntactic method for detection and correction of misrecognized mathematical formulae for a practical mathematical OCR system. Linear monadic context-free tree grammar (LM-CFTG) is employed as a formal framework to define syntactically acceptable mathematical formulae. For the purpose of practical evaluation, a verification system is developed, and the effectiveness of the method is demonstrated by using the ground-truthed mathematical document database InftyCDB-1 and a misrecognition database newly constructed for this study. A satisfactory number of misrecognitions are detected and delivered to the correction process..
265. Walaa Aly, Seiichi Uchida, and Masakazu Suzuki, Automatic Classification of Spatial Relationships among Mathematical Symbols Using Geometric Features, IEICE Trans. Information & Systems, vol. -D, 2009.11.
266. Toru Wakahara, Seiichi Uchida, Hierarchical Decomposition of Handwriting Deformation Vector Field Using 2D Warping and Global/Local Affine Transformation, Proceedings of The 10th International Conference on Document Analysis and Recognition, vol.J91-D, no.8, 2009.07.
267. Yoshinori Katayama, Seiichi Uchida and Hiroaki Sakoe,, Stochastic Model of Stroke Order Variation, Proceedings of The 10th International Conference on Document Analysis and Recognition, 2009.07.
268. Kazumasa Iwata, Koichi Kise, Tomohiro Nakai, Masakazu Iwamura, Seiichi Uchida and Shinichiro Omachi,, Capturing Digital Ink as Retrieving Fragments of Document Images, Proceedings of The 10th International Conference on Document Analysis and Recognition, vol.29, no.9, pp.1326-1332, 2009.07.
269. Akio Fujiyoshi, Masakazu Suzuki and Seiichi Uchida,, Syntactic Detection and Correction of Misrecognitions in Mathematical OCR, Proceedings of The 10th International Conference on Document Analysis and Recognition, 2009.07.
270. Seiichi Uchida, Ryoji Hattori, Masakazu Iwamura, Shinichiro Omachi and Koichi Kise,, Conspicuous character patterns, Proceedings of The 10th International Conference on Document Analysis and Recognition, vol.J91-D, no.5, pp.1434-1441, 2009.07.
271. Walaa Aly, Seiichi Uchida, Akio Fujiyoshi and Masakazu Suzuki,, Statistical classification of spatial relationships among mathematical symbols, Proceedings of The 10th International Conference on Document Analysis and Recognition, vol.J91-D, no.5, pp.1380-1392, 2009.07.
272. Koichi Kise, Kazumasa Iwata, Tomohiro Nakai, Masakazu Iwamura, Seiichi Uchida and Shinichiro Omachi, Document-Level Positioning of a Pen Tip by Retrieval of Image Fragments, Proceedings of The Third International Workshop on Camera-Based Document Analyais and Recognition, vol.41, no.4, pp.1230-1240, 2009.07.
273. Seiichi Uchida, Ryoji Hattori, Masakazu Iwamura, Shinichiro Omachi and Koichi Kise?, Selecting and Evaluating Conspicuous Character Patterns, Proceedings of The Third International Workshop on Camera-Based Document Analyais and Recognition, 25OF4-N52, 2009.07.
274. Seiichi Uchida, Katsuhiro Itou, Masakazu Iwamura, Shinichiro Omachi and Koichi Kise?, On a Possibility of Pen-Tip Camera for the Reconstruction of Handwritings, Proceedings of The Third International Workshop on Camera-Based Document Analyais and Recognition, vol.J91-D, no.1, pp.136-138, 2009.07.
275. Visual Tracking of an Object with its Motion Information.
276. Masakazu Iwamura, Ryo Niwa, Akira Horimatsu, Koichi Kise, Seiichi Uchida, Shinichiro Omachi, Layout-free dewarplng of planar document images, Document Recognition and Retrieval XVI Proceedings of SPIE - The International Society for Optical Engineering, 10.1117/12.806122, 7247, 2009.03, For user convenience, processing of document images captured by a digital camera has been attracted much attention. However, most existing processing methods require an upright image such like captured by a scanner. Therefore, we have to cancel perspective distortion of a camera-captured image before processing. Although there are rectification methods of the distortion, most of them work under certain assumptions on the layout; the borders of a document are available, lextlines are in parallel, a stereo camera or a video image is required and so on. In this paper, we propose a layout-free rectification method which requires none of the above assumptions. We confirm the effectiveness of the proposed method by experiments..
277. Masakazu Iwamura, Ryo Niwa, Akira Horimatsu, Koichi Kise, Seiichi Uchida, and Shinichiro Omachi, Layout-Free Dewarping of Planar Document Images, Document Recognition and Retrieval XVI, 2009.01.
278. Walaa Aly, Seiichi Uchida, Masakazu Suzuki, Automatic classification of spatial relationships among mathematical symbols using geometric features, IEICE Transactions on Information and Systems, 10.1587/transinf.E92.D.2235, E92-D, 11, 2235-2243, 2009.01, Machine recognition of mathematical expressions on printed documents is not trivial even when all the individual characters and symbols in an expression can be recognized correctly. In this paper, an automatic classification method of spatial relationships between the adjacent symbols in a pair is presented. This classification is important to realize an accurate structure analysis module of math OCR. Experimental results on very large databases showed that this classification worked well with an accuracy of 99.525% by using distribution maps which are defined by two geometric features, relative size and relative position, with careful treatment on document-dependent characteristics..
279. Atsutoshi Shimeno, Seiichi Uchida, Ryo Kurazume, Rin Ichiro Taniguchi, Tsutomu Hasegawa, Visual tracking of an object with its motion information, IEEJ Transactions on Electronics, Information and Systems, 10.1541/ieejeiss.129.977, 129, 5, 977-984+28, 2009.01, Tracking of a moving robot in surveillance video is an important task for coexistence of human beings with robots. An essential technology to manage coexistence environment of human beings and moving robots is separation and tracking of moving robots. For this task, the moving robot should be separated from other moving objects, i.e., human beings. We assume that the robot provides its additional motion information to the surveillance system to ease the task. The robot can be tracked from the other objects as a moving region being consistent with the additional motion information. For this purpose, we modify a tracking algorithm based on particle filter in order to incorporate the additional motion information. The results of an experiment on real surveillance video sequences have indicated that the proposed framework can separate and track a moving robot under the existence of several walking persons..
280. Koiehi Kise, Kazumasa Iwata, Tomohiro Nakai, Masakazu Iwamura, Seiichi Uchida, Shinichiro Omachi, Document-Level positioning of a pen tip by retrieval of image fragments, 3rd International Workshop on Camera-Based Document Analysis and Recognition, CBDAR 2009 Proceedings of the 3rd International Workshop on Camera-Based Document Analysis and Recognition, CBDAR 2009, 61-68, 2009, This paper presents a new method of positioning a pen tip on a document for pen-based computing. Current technologies such as the Anoto pens rely on special mechanisms for positioning, which result in limiting their applicability.Our proposal is to ease this problem - a camera pen that allows us to write on ordinary paper for positioning. A document image retrieval method calledLLAH is tuned to position the pen tip efficiently on the coordinates of a document only by capturing its tiny image fragment. In this paper we report results of preliminary experiments about recovering handwriting of simple figures and letters on documents using a prototype camera pen..
281. Seiichi Uchida, Katsuhiro Itou, Masakazu Iwamura, Shinichiro Omachi, Koichi Kise, On a possibility of pen-tip camera for the reconstruction of handwritings, 3rd International Workshop on Camera-Based Document Analysis and Recognition, CBDAR 2009 Proceedings of the 3rd International Workshop on Camera-Based Document Analysis and Recognition, CBDAR 2009, 119-126, 2009, Toward realization of "writing-life-log," a camera-based handwriting pattern acquisition system is proposed. The camera is attached around the tip of a popular pen and captures frame images around the pen tip continuously. Our problem is the reconstruction of the entire handwriting pattern by video-mosaicing of those frame images with perspective registration of consecutive frames. A key idea is to use microscopic structure of paper surface, called paper fingerprint, for the registration. Specifically, perspective transformation is estimated by using correspondence of SURF keypoints extracted on paper surface. Since the precise structure can be captured stably as the SURF keypoints from the pen-tip camera, thus it is possible to expect accurate registration of video frames..
282. Seiichi Uchida, Ryoji Hattori, Masakazu Iwamura, Shinichiro Omachi, Koichi Kise, Selecting and evaluating conspicuous character patterns, 3rd International Workshop on Camera-Based Document Analysis and Recognition, CBDAR 2009 Proceedings of the 3rd International Workshop on Camera-Based Document Analysis and Recognition, CBDAR 2009, 111-118, 2009, Character detection in scenery images is a very difficult task. This paper describes a strategy of selecting character patterns for easier detection in scenery image. These character patterns, called conspicuous character patterns, are selected from character font sets according to a criterion that evaluates how the font has a larger distance from a non-character pattern distribution and, simultaneously, with a smaller distance from a general character pattern distribution. Qualitative and quantitative evaluations have been made to observe how the selected fonts are more conspicuous than other fonts..
283. Yoshinori Katayama, Seiichi Uchida, Hiroaki Sakoe, Stochastic model of stroke order variation, ICDAR2009 - 10th International Conference on Document Analysis and Recognition ICDAR2009 - 10th International Conference on Document Analysis and Recognition, 10.1109/ICDAR.2009.146, 803-807, 2009, A stochastic model of stroke order variation is proposed and applied to the stroke-order free on-line Kanji character recognition. The proposed model is a hidden Markov model (HMM) with a special topology to represent all stroke order variations. A sequence of state transitions from the initial state to the final state of the model represents one stroke order and provides a probability of the stroke order. The distribution of the stroke order probability can be trained automatically by using an EM algorithm from a training set of on-line character patterns. Experimental results on large-scale test patterns showed that the proposed model could represent actual stroke order variations appropriately and improve recognition accuracy by penalizing incorrect stroke orders..
284. Seiichi Uchida, Kazuma Amamoto,, Early Recognition of Sequential Patterns by Classifier Combination, Proceedings of 19th IAPR International Conference on Pattern Recognition (ICPR 2008), vol.7, no.4, pp.709-733, Oct. 2007., 2008.12.
285. Yoshinori Katayama, Seiichi Uchida, and Hiroaki Sakoe,, A New HMM for On-Line Character Recognition Using Pen-Direction and Pen-Coordinate Features, Proceedings of 19th IAPR International Conference on Pattern Recognition (ICPR 2008), 2008.12.
286. Walaa Aly, Seiichi Uchida, and Masakazu Suzuki, Identifying subscripts and superscripts in mathematical documents, Mathematics in Computer Science, 2008.12.
287. Walaa Aly, Seiichi Uchida, Masakazu Suzuki, A large-scale analysis of mathematical expressions for an accurate understanding of their structure, 8th IAPR International Workshop on Document Analysis Systems, DAS 2008 DAS 2008 - Proceedings of the 8th IAPR International Workshop on Document Analysis Systems, 10.1109/DAS.2008.53, 549-556, 2008.12, A wide variety of mathematical expressions printed in scientific and technical reports can be recognized by analyzing the two-dimensional layout structure. In this paper, the position relation between adjacent characters is analyzed for the purpose of automatic discrimination between baseline, subscript, and superscript characters. This analyzing is one of the most important parts of structure analysis. The proposed method is very promising, as the results reached up to (99.76%) over a very large database by using distribution map. This distribution map is defined by two important features, i.e., relative size and relative position..
288. Yoshinori Katayama, Seiichi Uchida, Hiroaki Sakoe, A new HMM for on-line character recognition using pen-direction and pen-coordinate features, 2008 19th International Conference on Pattern Recognition, ICPR 2008 2008 19th International Conference on Pattern Recognition, ICPR 2008, 2008.12, A new hidden Markov model (HMM) is proposed for on-line character recognition using two typical features, pen-direction feature and pen-coordinate feature. These two features are quite different in their stationarity; pen-direction feature is stationary within every line segment of a stroke whereas pen-coordinate feature is not. In the proposed HMM, these contrasting features are used in a separative and selective way. Specifically speaking, pen-direction feature is outputted repeatedly at intra-state transition whereas pen-coordinate feature is outputted once at inter-state transition. The superiority of the proposed HMM over the conventional HMMs was shown through single-stroke and multi-stroke character recognition experiments..
289. Akira Horimatsu, Ryo Niwa, Seiichi Uchida, Masakazu Iwamura, Shinichiro Omachi, Koichi Kise, Affine invariant recognition of characters by progressive pruning, 8th IAPR International Workshop on Document Analysis Systems, DAS 2008 DAS 2008 - Proceedings of the 8th IAPR International Workshop on Document Analysis Systems, 10.1109/DAS.2008.88, 237-244, 2008.12, There are many problems to realize camera-based character recognition. One of the problems is that characters in scenes are often distorted by geometric transformations such as affine distortions. Although some methods that remove the affine distortions have been proposed, they cannot remove a rotation transformation of a character. Thus a skew angle of a character has to be determined by examining all the possible angles. However, this consumes quite a bit of time. In this paper, in order to reduce the processing time for an affine invariant recognition, we propose a set of affine invariant features and a new recognition scheme called "progressive pruning." The progressive pruning gradually prunes less feasible categories and skew angles using multiple classifiers. We confirmed the progressive pruning with the affine invariant features reduced the processing time at least less than half without decreasing the recognition rate..
290. Seiichi Uchida, Elastic matching techniques for handwritten character recognition, Pattern Recognition Technologies and Applications Recent Advances, 10.4018/978-1-59904-807-9.ch002, 17-38, 2008.12, This chapter reviews various elastic matching techniques for Handwritten character recognition. Elastic matching is formulated as an optimization problem of planar matching, or pixel-to-pixel correspondence, between two character images under a certain matching model, such as affine and nonlinear. Use of elastic matching instead of rigid matching improves the robustness of recognition systems against geometric deformations in Handwritten character images. In addition, the optimized matching represents the deformation of Handwritten characters and thus is useful for statistical analysis of the deformation. This chapter argues the general property of elastic matching techniques and their classification by matching models and optimization strategies. It also argues various topics and future work related to elastic matching for emphasizing theoretical and practical importance of elastic matching..
291. Walaa Aly, Seiichi Uchida, Masakazu Suzuki, Identifying subscripts and superscripts in mathematical documents, Mathematics in Computer Science, 10.1007/s11786-008-0051-9, 2, 2, 195-209, 2008.12, In mathematical OCR, it is necessary to analyze two-dimensional structures of the component characters and symbols in mathematical expressions printed in scientific documents. In this paper, we analyze the positional relationships between adjacent characters for the purpose of automatic discrimination between baseline characters, subscripts, and superscripts, which is one of the most important and delicate parts of structure analysis. It has been proven through a large-scale experiment that this discrimination can be carried out almost perfectly (~99.89%) by using the relative size and position of adjacent characters..
292. Kenichi Morooka, Xian Chen, Ryo Kurazume, Seiichi Uchida, Hara Kenji, Yumi Iwashita, Makoto Hashizume, Real-time nonlinear FEM with neural network for simulating soft organ model deformation, 11th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2008 Medical Image Computing and Computer-Assisted Intervention - MICCAI 2008 - 11th International Conference, Proceedings, 10.1007/978-3-540-85990-1-89, 742-749, 2008.12, This paper presents a new method for simulating the deformation of organ models by using a neural network. The proposed method is based on the idea proposed by Chen et al. [2] that a deformed model can be estimated from the superposition of basic deformation modes. The neural network finds a relationship between external forces and the models deformed by the forces. The experimental results show that the trained network can achieve a real-time simulation while keeping the acceptable accuracy compared with the nonlinear FEM computation..
293. Akihiro Mori and Seiichi Uchida, Fast image mosaicing based on histograms, IEICE Transactions on Information and Systems, 2008.11.
294. Akihiro Mori, Seiichi Uchida, Fast image mosaicing based on histograms, IEICE Transactions on Information and Systems, 10.1093/ietisy/e91-d.11.2701, E91-D, 11, 2701-2708, 2008.11, This paper introduces a fast image mosaicing technique that does not require costly search on image domain (e.g., pixel-to-pixel correspondence search on the image domain) and the iterative optimization (e.g., gradient-based optimization, iterative optimization, and random optimization) of geometric transformation parameter. The proposed technique is organized in a two-step manner. At both steps, histograms are fully utilized for high computational efficiency. At the first step, a histogram of pixel feature values is utilized to detect pairs of pixels with the same rare feature values as candidates of corresponding pixel pairs. At the second step, a histogram of transformation parameter values is utilized to determine the most reliable transformation parameter value. Experimental results showed that the proposed technique can provide reasonable mosaicing results in most cases with very feasible computations..
295. Seiichi Uchida, Megumi Sakai, Masakazu Iwamura, Shinichiro Omachi, and Koichi Kise,, Skew Estimation by Instances, Proceedings of The Eighth International Workshop on Document Analysis Systems, 2008.09.
296. Akira Horimatsu, Ryo Niwa, Masakazu Iwamura, Koichi Kise, Seiichi Uchida, and Shinichiro Omachi,, Affine Invariant Recognition of Characters by Progressive Pruning, Proceedings of The Eighth International Workshop on Document Analysis Systems, 2008.09.
297. Walaa Aly, Seiichi Uchida, and Masakazu Suzuki,, A Large-Scale Analysis of Mathematical Expressions for an Accurate Understanding of Their Structure, Proceedings of The Eighth International Workshop on Document Analysis Systems, 2008.09.
298. Ken'ichi Morooka, Xian Chen, Ryo Kurazume, Uchida Seiichi, Kenji Hara, Yumi Iwashita, Makoto Hashizume, Real-time Nonlinear FEM with Neural Network for Simulating Soft Organ Model Deformation, Proceedings of The 11th International Conference on Medical Image Computing and Computer Assisted Intervention, 2008.09.
299. Yumi Iwashita, Ryo Kurazume, Kenji Hara, Seiichi Uchida, Ken'ichi Morooka, Tsutomu Hasegawa, Fast 3D reconstruction of human shape and motion tracking by parallel fast level set method, 2008 IEEE International Conference on Robotics and Automation, ICRA 2008 2008 IEEE International Conference on Robotics and Automation, ICRA 2008, 10.1109/ROBOT.2008.4543332, 980-986, 2008.09, This paper presents a parallel algorithm of the Level Set Method named the Parallel Fast Level Set Method, and its application for real-time 3D reconstruction of human shape and motion. The Fast Level Set Method is an efficient implementation algorithm of the Level Set Method and has been applied to several applications such as object tracking in video images and 3D shape reconstruction using multiple stereo cameras. In this paper, we implement the Fast Level Set Method on a PC cluster and develop a real-time motion capture system for arbitrary viewpoint image synthesis. To obtain high performance on a PC cluster, efficient load-balancing and resource allocation algorithms are crucial problems. We develop a novel optimization technique of load distribution based on the estimation of moving direction of object boundaries. In this technique, the boundary motion is estimated in the framework of the Fast Level Set Method, and the optimum load distribution is predicted and performed according to the estimated boundary motion and the current load balance. Experiments of human shape reconstruction and arbitrary viewpoint image synthesis using the proposed system are successfully carried out..
300. Akio Fujiyoshi, Masakazu Suzuki, Seiichi Uchida, Verification of mathematical formulae based on a combination of context-free grammar and tree grammar, 9th Int. Conf. Artificial Intelligence and Symbolic Computation, AISC 2008 - 15th Symposium on the Integration of Symbolic Computation and Mechanized Reasoning, Calculemus 2008 - 7th Int. Conf. Mathematical Knowledge Management, MKM 2008 Intelligent Computer Mathematics - 9th International Conference, AISC 2008 - 15th Symposium, Calculemus 2008 - 7th International Conference, MKM 2008, Proceedings, 10.1007/978-3-540-85110-3_35, 415-429, 2008.09, This paper proposes the use of a formal grammar for the verification of mathematical formulae for a practical mathematical OCR system. Like a C compiler detecting syntax errors in a source file, we want to have a verification mechanism to find errors in the output of mathematical OCR. Linear monadic context-free tree grammar (LM-CFTG) was employed as a formal framework to define "well-formed" mathematical formulae. For the purpose of practical evaluation, a verification system for mathematical OCR was developed, and the effectiveness of the system was demonstrated by using the ground-truthed mathematical document database INFTY CDB-1..
301. Seiichi Uchida, Kazuya Niyagawa, Hiroaki Sakoe,, Feature Desynchronization in Online Character Recognition, Proceedings of the 11th International Conference on Frontiers of Handwriting Recognition, 2008.08.
302. Akio Fujiyoshi, Masakazu Suzuki, Seiichi Uchida,, Verification of Mathematical Formulae Based on a Combination of Context-Free Grammar and Tree Grammar, Proceedings of The Seventh International Conference on Mathematical Knowledge Management, 2008.07.
303. Christopher D. Malon, Seiichi Uchida, Masakazu Suzuki, Mathematical symbol recognition with support vector machines, Pattern Recognition Letters, 2008.07.
304. Christopher Malon, Seiichi Uchida, Masakazu Suzuki, Mathematical symbol recognition with support vector machines, Pattern Recognition Letters, 10.1016/j.patrec.2008.02.005, 29, 9, 1326-1332, 2008.07, Single-character recognition of mathematical symbols poses challenges from its two-dimensional pattern, the variety of similar symbols that must be recognized distinctly, the imbalance and paucity of training data available, and the impossibility of final verification through spell check. We investigate the use of support vector machines to improve the classification of InftyReader, a free system for the OCR of mathematical documents. First, we compare the performance of SVM kernels and feature definitions on pairs of letters that InftyReader usually confuses. Second, we describe a successful approach to multi-class classification with SVM, utilizing the ranking of alternatives within InftyReader's confusion clusters. The inclusion of our technique in InftyReader reduces its misrecognition rate by 41%..
305. Yumi Iwashita, Ryo Kurazume, Kenji Hara, Seiichi Uchida, Ken'ichi Morooka, Tsutomu Hasegawa,, Fast 3D Reconstruction of Human Shape and Motion Tracking by Parallel Fast Level Set Method, Proceedings of 2008 IEEE International Conference on Robotics and Automation, 2008.05.
306. 2D/3D Registration by Back Projection and Geometrical Constraints.
307. Seiichi Uchida, Hiromitsu Miyazaki, Hiroaki Sakoe, Mosaicing-by-recognition for video-based text recognition, Pattern Recognition, vol.24, no.8, pp.954--963, 2008.04.
308. Seiichi Uchida, Hiromitsu Miyazaki, Hiroaki Sakoe, Mosaicing-by-recognition for video-based text recognition, Pattern Recognition, 10.1016/j.patcog.2007.08.005, 41, 4, 1230-1240, 2008.04, Text recognition captured in multiple frames by a hand-held video camera is a challenging task because it is possible to capture and recognize a longer line of text while improving the quality of the text image by utilizing the redundancy of the overlapping areas between the frames. For this task, the video frames should be registered, i.e., mosaiced, after compensating for their distortions due to camera shakes. In this paper, a mosaicing-by-recognition technique is proposed where the problems of video mosaicing and text recognition are formulated as a unified optimization problem and solved by a dynamic programming-based optimization algorithm simultaneously and collaboratively. Experimental results indicate that, even if the frames undergo various distortions such as rotation, scaling, translation, and nonlinear speed fluctuation of camera movement, the proposed technique provides fine mosaic image by accurate distortion estimation (around 90% of perfect estimation) and character recognition accuracy (over 95%)..
309. Yuji Shinomura, Tomotaka Harano, Toru Tamaki, Toshiyuki Amano, Kazufumi Kaneda, Seiichi Uchida,, Comparative study of path nomalizations for path prediction, Proceedings of 14th Korea-Japan Joint Workshop on Frontiers of Computer Vision, pp.61-66, 2008.01.
310. Kenichi Morooka, Xian Chen, Ryo Kurazume, Seiichi Uchida, Hara Kenji, Yumi Iwashita, Makoto Hashizume, Real-time nonlinear FEM with neural network for simulating soft organ model deformation., Medical image computing and computer-assisted intervention : MICCAI ... International Conference on Medical Image Computing and Computer-Assisted Intervention, 11, Pt 2, 742-749, 2008.01, This paper presents a new method for simulating the deformation of organ models by using a neural network. The proposed method is based on the idea proposed by Chen et al. that a deformed model can be estimated from the superposition of basic deformation modes. The neural network finds a relationship between external forces and the models deformed by the forces. The experimental results show that the trained network can achieve a real-time simulation while keeping the acceptable accuracy compared with the nonlinear FEM computation..
311. Seiichi Uchida, Kazuma Amamoto, Early recognition of sequential patterns by classifier combination, 2008 19th International Conference on Pattern Recognition, ICPR 2008 2008 19th International Conference on Pattern Recognition, ICPR 2008, 2008, This paper proposes an early recognition method, i.e., a method for recognizing sequential patterns at their beginning parts. The method is based on a combination of frame classifiers prepared at individual frames. The training patterns misrecognized by the frame classifier at a certain frame are heavily weighted for the complementary training of the frame classifier at the next frame. The method was applied to an online character recognition task for showing its usefulness..
312. Seiichi Uchida, Megumi Sakai, Masakazu Iwamura, Shinichiro Omachi, Koichi Kise, Extraction of embedded class information from universal character pattern, 9th International Conference on Document Analysis and Recognition, ICDAR 2007 Proceedings - 9th International Conference on Document Analysis and Recognition, ICDAR 2007, 10.1109/ICDAR.2007.4378747, 437-441, 2007.12, This paper is concerned with a universal pattern, which is defined as a character pattern designed to have high machine-readability. This universal pattern is a character pattern printed with stripes. The cross ratio calculated from the widths of the stripes represents the character class. Thus, if the boundaries of the stripes can be detected for measuring the widths, the class can be determined without ordinary recognition process. Furthermore, since the cross ratio is invariant to projective distortions, the correct class will be still determined under those distortions. This paper describes a practical scheme to recognize this universal pattern. The proposed scheme includes a novel algorithm to detect the stripe boundaries stably even from the universal pattern image contaminated by non-uniform lighting and noise. The algorithm is realized by a combination of a dynamic programming-based optimal boundary detection and a finite state automaton which represents the property of the universal pattern. Experimental results showed the proposed scheme could recognize 99.6% of the universal pattern images which underwent heavy projective distortions and non-uniform lighting..
313. V. Bucha, S. Uchida, S. Ablameyko, Image pixel force fields and their application for color map vectorisation, 9th International Conference on Document Analysis and Recognition, ICDAR 2007 Proceedings - 9th International Conference on Document Analysis and Recognition, ICDAR 2007, 10.1109/ICDAR.2007.4377111, 1228-1232, 2007.12, Pixel force field is a novel image representation where at each pixel a two-dimensional vector is defined for representing the circumstance of the pixel. The vector is oriented to the center of the region composed of vectors having the same qualitative property, such as color and gray-scale level. Using the pixel force field, that is, the orientation and the magnitude of the vector, many fundamental and specific image processing tasks can be solved. As examples of the tasks, the force field is applied to color image thinning, color image segmentation, and color map vectorisation..
314. Seiichi Uchida, Akihiro Mori, Ryo Kurazume, Rin Ichiro Taniguchi, Tsutomu Hasegawa, Logical DP matching for detecting similar subsequence, 8th Asian Conference on Computer Vision, ACCV 2007 Computer Vision - ACCV 2007 - 8th Asian Conference on Computer Vision, Proceedings, 628-637, 2007.12, A logical dynamic programming (DP) matching algorithm is proposed for extracting similar subpatterns from two sequential patterns. In the proposed algorithm, local similarity between two patterns is measured by a logical function, called support. The DP matching with the support can extract all similar subpatterns simultaneously while compensating nonlinear fluctuation. The performance of the proposed algorithm was evaluated qualitatively and quantitatively via an experiment of extracting motion primitives, i.e., common subpatterns in gesture patterns of different classes..
315. Roman Bertolami, Seiichi Uchida, Matthias Zimmermann, Horst Bunke, Non-uniform slant correction for handwritten text line recognition, 9th International Conference on Document Analysis and Recognition, ICDAR 2007 Proceedings - 9th International Conference on Document Analysis and Recognition, ICDAR 2007, 10.1109/ICDAR.2007.4378668, 18-22, 2007.12, In this paper we apply a novel non-uniform slant correction preprocessing technique to improve the recognition of offline handwritten text lines. The local slant correction is expressed as a global optimisation problem of the sequence of local slant angles. This is different to conventional slant removal techniques that rely on the average slant angle. Experiments based on a state-of-the-art handwritten text line recogniser show a significant gain in word level accuracy for the investigated preprocessing methods..
316. Daiki Baba, Seiichi Uchida, Hiroaki Sakoe, Predictive DP matching for on-line character recognition, 9th International Conference on Document Analysis and Recognition, ICDAR 2007 Proceedings - 9th International Conference on Document Analysis and Recognition, ICDAR 2007, 10.1109/ICDAR.2007.4377000, 674-678, 2007.12, For on-line character recognition, predictive DP matching is proposed where two physically different features, coordinate features and directional features, are handled in a unified manner. For this unification, the distance of the directional features is converted into a distance of the coordinate features by a feature prediction technique. An experimental result showed that the predictive DP matching could attain a recognition rate comparable to the rate by the conventional DP matching which requires the costly optimization of the weight to balance the two features..
317. Seiichi Uchida, Akihiro Mori, Ryo Kurazume, Rin-ichiro Taniguchi, Tsutomu Hasegawa,, Logical DP Matching for Detecting Similar Subsequence, Proceedings of 8th Asian Conference on Computer Vision, 2007.11.
318. Sergey V. Ablameyko, Seiichi Uchida, Recognition of engineering drawing entities: review of approaches, International Journal of Image and Graphics, 2007.10.
319. Sergey V. Ablameyko, Seiichi Uchida, Recognition of engineering drawing entities
Review of approaches, International Journal of Image and Graphics, 10.1142/S0219467807002878, 7, 4, 709-733, 2007.10, Recognition of engineering drawing entities is one of the most difficult stages in engineering drawing interpretation and many attempts to recognize various types of ED entities have been made. In this paper, we review algorithms for the recognition of ED entities, especially dimensions and crosshatching areas. For the recognition of dimensions, we analyze how dimension texts can be separated from graphics and how arrowheads of dimension lines are recognized. We also analyze the recent systems of ED interpretation. Finally, future tasks are discussed..
320. Victor Bucha, Sergey Ablameyko, Seiichi Uchida, Image pixel force fields and their application for color map vectorisation, Proceedings of 9th International Conference on Document Analysis and Recognition, 2007.09.
321. Daiki Baba, Seiichi Uchida, Hiroaki Sakoe, Predictive DP Matching for On-Line Character Recognition, Proceedings of 9th International Conference on Document Analysis and Recognition, 2007.09.
322. Seiichi Uchida, Megumi Sakai, Masakazu Iwamura, Shinichiro Omachi, Koichi Kise, Extraction of Embedded Class Information from Universal Character Pattern, Proceedings of 9th International Conference on Document Analysis and Recognition, 2007.09, This paper is concerned with a universal pattern, which
is defined as a character pattern designed to have high
machine-readability. This universal pattern is a character
pattern printed with stripes. The cross ratio calculated
from the widths of the stripes represents the character class.
Thus, if the boundaries of the stripes can be detected for
measuring the widths, the class can be determined without
ordinary recognition process. Furthermore, since the cross
ratio is invariant to projective distortions, the correct class
will be still determined under those distortions. This paper
describes a practical scheme to recognize this universal
pattern. The proposed scheme includes a novel algorithm to
detect the stripe boundaries stably even from the universal
pattern image contaminated by non-uniform lighting and
noise. The algorithm is realized by a combination of a dynamic
programming-based optimal boundary detection and
a finite state automaton which represents the property of
the universal pattern. Experimental results showed the proposed
scheme could recognize 99.6% of the universal pattern
images which underwent heavy projective distortions
and non-uniform lighting..
323. Roman Bertolami, Seiichi Uchida, Matthias Zimmermann, Horst Bunke,, Non-Uniform Slant Correction for Handwritten Text Line Recognition, Proceedings of 9th International Conference on Document Analysis and Recognition, 2007.09.
324. Masakazu Iwamura, Ryo Niwa, Koichi Kise, Seiichi Uchida, Shinichiro Omachi,, Rectifying Perspective Distortion into Affine Distortion Using Variants and Invariants, Proceedings of Second International Workshop on Camera-Based Document Analysis and Recognition 2007, 2007.09.
325. Seiichi Uchida, Megumi Sakai, Masakazu Iwamura, Shinichiro Omachi, Koichi Kise,, Instance-Based Skew Estimation of Document Images by a Combination of Variant and Invariant, Proceedings of Second International Workshop on Camera-Based Document Analysis and Recognition 2007, 2007.09.
326. Ryoji Hattori, Seiichi Uchida, Color quantization for scene change detection, Proceedings of The First International Symposium on Information and Computer Elements, 2007.09.
327. Atsutoshi Shimeno, Seiichi Uchida, Ryo Kurazume, Rin-ichiro Taniguchi, Tsutomu Hasegawa, Separation and tracking of moving object using rough motion information from the object, Proceedings of The First International Symposium on Information and Computer Elements, Document Analysis Systems VII, Lecture Notes in Computer Sciences 3872, pp.153-163, 2006, 2007.09.
328. Ken'ichi Morooka, Hiroshi Masuda, Ryo Kurazume, Xian Chen, Seiichi Uchida, Kenji Hara, Makoto Hashizume, Real time estimation of deforming organs by neural network for endoscopic surgery simulator, Proceedings of The First International Symposium on Information and Computer Elements, vol.J89-D, no.2, pp.344--352, 2007.09.
329. Fast 3D Shape Reconstruction of Moving Object by Parallel Fast Level Set Method.
330. Detection of Similar Sub-Sequence by Logical DP Matching.
331. Masakazu Suzuki, Christopher Malon, Seiichi Uchida, Databases of mathematical documents, Research Reports on Information Science and Electrical Engineering of Kyushu University, 302-306, 2007.04.
332. Seiichi Uchida, Megumi Sakai, Masakazu Iwamura, Shinichiro Omachi, Koichi Kise, Instance-Based skew estimation of document images by a combination of variant and invariant, 2nd International Workshop on Camera-Based Document Analysis and Recognition, CBDAR 2007 Proceedings of the 2nd International Workshop on Camera-Based Document Analysis and Recognition, CBDAR 2007, 53-60, 2007, A novel technique for estimating geometric deformations is proposed and applied to document skew (i.e., rotation) estimation. The proposed method possesses two novel properties. First, the proposed method estimates the skew angles at individual connected components. Those skew angles are then voted to determine the skew angle of the entire document. Second, the proposed method is based on instancebased learning. Specifically, a rotation variant and a rotation invariant are learned, i.e., stored as instances for each character category, and referred for estimating the skew angle very efficiently. The result of a skew estimation experiment on 55 document images has shown that the skew angles of 54 document images were successfully estimated with errors smaller than 2.0 degree. The extension for estimating perspective deformation is also discussed for the application to camera-based OCR..
333. Masakazu Iwamura, Ryo Niwa, Koichi Kise, Seiichi Uchida, Shinichiro Omachi, Rectifying perspective distortion into affine distortion using variants and invariants, 2nd International Workshop on Camera-Based Document Analysis and Recognition, CBDAR 2007 Proceedings of the 2nd International Workshop on Camera-Based Document Analysis and Recognition, CBDAR 2007, 138-145, 2007, For user convenience, document image processing captured with a digital camera instead of a scanner has been researched. However, existing methods of document image processing are not usable for a perspective document image captured by a digital camera because most of them are designed for the one captured by a scanner. Thus, we have to rectify the perspective of the document image and obtain the frontal image as if it was captured by a scanner. In this paper, for eliminating perspective distortion from a planar paper without any prior knowledge, we propose a new rectification method of a document image introducing variants which change according to the gradient of the paper and invariants which do not change against it. Since the proposed method does not use strong assumptions, it is widely applicable to many document images unlike other methods. We confirmed the proposed method rectifies a document image suffering from perspective distortion and acquires the one with affine distortion..
334. Shinichiro Omachi, Seiichi Uchida, Masakazu Iwamura, Koichi Kise, Affine invariant information embedment for accurate camera-based character recognition, 18th International Conference on Pattern Recognition, ICPR 2006 Proceedings - 18th International Conference on Pattern Recognition, ICPR 2006, 10.1109/ICPR.2006.229, 1098-1101, 2006.12, Recognizing characters in a scene image taken by a digital camera has been studied for decades. However, it is still a challenging problem to achieve high accuracy. In this paper, we propose a method of embedding information in a character pattern so that the class of the character can be identified. The information should be robust against geometric distortions since an image taken by a digital camera is usually geometrically distorted. In the proposed method, a character pattern is designed in two colors so that the information is embedded as the area ratio of regions of two colors. Since the area ratio is affine invariant, it is expected that the area ratio is correctly extracted even if a character image is affine-transformed. We generate character patterns with the embedded information and discuss the effectiveness of the proposed method..
335. Wenjie Cai, Seiichi Uchida, Hiroaki Sakoe, An efficient radical-based algorithm for stroke-order-free online kanji character recognition, 18th International Conference on Pattern Recognition, ICPR 2006 Proceedings - 18th International Conference on Pattern Recognition, ICPR 2006, 10.1109/ICPR.2006.241, 986-989, 2006.12, This paper investigates improvements of an online handwriting stroke-order analysis algorithm -cube search, based on cube graph stroke-order generation model and dynamic programming (DP). By dividing character into radicals, the model is decomposed into infra-radical graphs and an inter-radical graph. This decomposition considerably reduces the time complexity of stroke-order search DP. Experimental results showed an significant improvements in operational speed. Additionally, recognition accuracy was also improved by prohibiting unnatural stroke-order..
336. Akihiro Mori, Seiichi Uchida, Ryo Kurazume, Rin Ichiro Taniguchi, Tsutomu Hasegawa, Hiroaki Sakoe, Early recognition and prediction of gestures, 18th International Conference on Pattern Recognition, ICPR 2006 Proceedings - 18th International Conference on Pattern Recognition, ICPR 2006, 10.1109/ICPR.2006.467, 560-563, 2006.12, This paper is concerned with an early recognition and prediction algorithm of gestures. Early recognition is the algorithm to provide recognition results before input gestures are completed. Motion prediction is the algorithm to predict the subsequent posture of the performer by using early recognition. In addition to them, this paper considers a gesture network for improving the performance of these algorithms. The performance of the proposed algorithm was evaluated by experiments of real-time control of a humanoid by gestures..
337. Ryo Kurazume, Hiroaki Omasa, Seiichi Uchida, Rinichiro Taniguchi, Tsutomu Hasegawa, Embodied proactive human interface "PICO-2", 18th International Conference on Pattern Recognition, ICPR 2006 Proceedings - 18th International Conference on Pattern Recognition, ICPR 2006, 10.1109/ICPR.2006.488, 1233-1237, 2006.12, We are conducting research on "Embodied Proactive Human Interface". The aim of this research is to develop a new human-friendly active interface based on two key technologies, an estimation mechanism of human intention for supporting natural communication named "Proactive Interface", and a tangible device using robot technology. This paper introduces the humanoid-type Two-legged robot named "PICO-2 ", which was developed as a tangible telecommunication device for the proactive human interface. In order to achieve the embodied telecommunication with PICO-2, we propose new tracking technique of human gestures using a monocular video camera mounted on PICO-2, and natural gesture reproduction by PICO-2 which absorbs the difference of body structure between the user and the robot..
338. A. Nedzved, S. Uchida, S. Ablameyko, Gray-scale thinning by using a pseudo-distance map, 18th International Conference on Pattern Recognition, ICPR 2006 Proceedings - 18th International Conference on Pattern Recognition, ICPR 2006, 10.1109/ICPR.2006.618, 239-242, 2006.12, In this paper, the algorithm for thinning of grey-scale images is proposed that is based on a pseudo-distance map (PDM). The PDM is a simplified distance map of gray-scale image and uses only that features of image and objects that are necessary to build a skeleton. The algorithm works fast for large gray-scale images and allows constructing a high quality skeleton..
339. Seiichi Uchida, Masakazu Iwamura, Shinichiro Omachi, Koichi Kise, OCR fonts revisited for camera-based character recognition, 18th International Conference on Pattern Recognition, ICPR 2006 Proceedings - 18th International Conference on Pattern Recognition, ICPR 2006, 10.1109/ICPR.2006.891, 1134-1137, 2006.12, In order to realize accurate camera-based character recognition, machine-readable class information is embedded into each character image. Specifically, each character image is printed with a pattern which comprises five stripes and the cross ratio derived from the pattern represents class information. Since the cross ratio is a projective invariant, the class information is extracted correctly regardless of camera angle. The results of simulation experiments showed that recognition rates over 99% were obtained by the extracted cross ratio under heavy projective distortions..
340. Masato Nakajima, Seiichi Uchida, Akihiro Mori, Ryo Kurazume, Rin-ichiro Taniguchi, Tsutomu Hasegawa, Hiroaki Sakoe, Motion prediction based on eigen-gestures, Proceedings of the 1st First Korea-Japan Joint Workshop on Pattern Recognition, 904-908, 2006.11.
341. Masakazu Iwamura, Yoshio Furuya, Koichi Kise, Shinichiro Omachi, Seiichi Uchida, Better decision boundary for pattern recognition with supplementary information, Proceedings of the 1st First Korea-Japan Joint Workshop on Pattern Recognition, E88D, 8, 1781-1790, 2006.11.
342. Early Recognition and Prediction of Gestures for Embodied Proactive Human Interface.
343. Seiichi Uchida, Kazuhiro Tanaka, Masakazu Iwamura, Shinichiro Omachi, and Koichi Kise, A data-embedding pen, Proceedings of the 10th International Workshop on Frontiers of Handwriting Recognition (IWFHR-10), 2006.10.
344. Y.Araki, D.Arita, R.Taniguchi, S.Uchida, R.Kurazume and T.Hasegawa, Construction of symbolic representation from human motion information, 10th Int. Conf. on Knowledge-Based & Intelligent Information & Engineering Systems, 2006.10.
345. Shinichiro Omachi, Masakazu Iwamura, Seiichi Uchida, and Koichi Kis, Affine invariant information embedment for accurate camera-based character recognition, Proceedings of 18th IAPR International Conference on Pattern Recognition (ICPR 2006), vol. 36, no. 5, pp. 13-22, 2006.08.
346. Seiichi Uchida, Masakazu Iwamura, Shinichiro Omachi, and Koichi Kise, OCR fonts revisited for camera-based character recognition, Proceedings of 18th IAPR International Conference on Pattern Recognition (ICPR 2006), 2006.08.
347. Ryo Kurazume, Hiroaki Omasa, Seiichi Uchida, Rin-ichiro Taniguchi, and Tsutomu Hasegawa, Embodied Proactive Human Interface ''PICO-2'', Proceedings of 18th IAPR International Conference on Pattern Recognition (ICPR 2006), pp.80-85, 2006.08.
348. Akihiro Mori, Seiichi Uchida, Ryo Kurazume, Rin-ichiro Taniguchi, Tsutomu Hasegawa and Hiroaki Sakoe, Early recognition and prediction of gestures, Proceedings of 18th IAPR International Conference on Pattern Recognition (ICPR 2006), 情報処理学会九州支部 推薦論文, 2006.08.
349. Wenjie Cai, Seiichi Uchida, Hiroaki Sakoe, An efficient radical-based algorithm for stroke-order-free online Kanji character recognition, Proceedings of 18th IAPR International Conference on Pattern Recognition (ICPR 2006), 3-8, 2006.08.
350. V.Bucha, S.Uchida, S.Ablameyko, Interactive Road Extraction with Pixel Force Fields, Proceedings of 18th IAPR International Conference on Pattern Recognition (ICPR 2006), 2006.08.
351. A.Nedzved, S.Uchida, S.Ablameyko, Gray-scale thinning by using a pseudo-distance map, Proceedings of 18th IAPR International Conference on Pattern Recognition (ICPR 2006), 224-227, 2006.08.
352. Christopher D. Malon, Seiichi Uchida, and Masakazu Suzuki, Support Vector Machines for Mathematical Symbol Recognition, 6th International Workshop on Statistical Pattern Recognition, 2006.08.
353. Seiichi Toyota, Seiichi Uchida, Masakazu Suzuki, Structural analysis of mathematical formulae with verification based on formula Description Grammar, 7th International Workshop on Document Analysis Systems, DAS 2006 Document Analysis Systems VII - 7th International Workshop, DAS 2006, Proceedings, 10.1007/11669487_14, 153-163, 2006.07, In this paper, a reliable and efficient structural analysis method for mathematical formulae is proposed for practical mathematical OCR. The proposed method consists of three steps. In the first step, a fast structural analysis algorithm is performed on each mathematical formula to obtain a tree representation of the formula. This step generally provides a correct tree representation but sometimes provides an erroneous representation. Therefore, the tree representation is verified by the following two steps. In the second step, the result of the analysis step, (i.e., a tree representation) is converted into a one-dimensional representation. The third step is a verification step where the one-dimensional representation is parsed by a formula description grammar, which is a context-free grammar specialized for mathematical formulae. If the one-dimensional representation is not accepted by the grammar, the result of the analysis step is detected as an erroneous result and alarmed to OCR users. This three-step organization achieves reliable and efficient structural analysis without any two-dimensional grammars..
354. Seiichi Toyota, Seiichi Uchida, and Masakazu Suzuki, Structural Analysis of Mathematical Formulae with Verification Based on Formula Description Grammar, Proceedings of 7th IAPR Workshop on Document Analysis Systems, 2006.02.
355. Yutaka Araki, Daisaku Arita, Rin-Ichiro Taniguchi, Seiichi Uchida, Ryo Kurazume, Tsutomu Hasegawa, Construction of symbolic representation from human motion information, 10th International Conference on Knowledge-Based Intelligent Information and Engineering Systems, KES 2006 Knowledge-Based Intelligent Information and Engineering Systems - 10th International Conference, KES 2006, Proceedings, 212-219, 2006.01, In general, avatar-based communication has a merit that it can represent non-verbal information. The simplest way of representing the non-verbal information is to capture the human action/motion by a motion capture system and to visualize the received motion data through the avatar. However, transferring raw motion data often makes the avatar motion unnatural or unrealistic because the body structure of the avatar is usually a bit different from that of the human beings. We think this can be solved by transferring the meaning of motion, instead of the raw motion data, and by properly visualizing the meaning depending on characteristics of the avatar's function and body structure. Here, the key issue is how to symbolize the motion meanings. Particularly, the problem is what kind of motions we should symbolize. In this paper, we introduce an algorithm to decide the symbols to be recognized referring to accumulated communication data, i.e., motion data..
356. Christopher Malon, Seiichi Uchida, Masakazu Suzuki, Support vector machines for mathematical symbol recognition, Joint IAPR International Workshops on Structural, Syntactic, and Statistical Pattern Recognition, SSPR 2006 and SPR 2006 Structural, Syntactic, and Statistical Pattern Recognition - Joint IAPR International Workshops, SSPR 2006 and SPR 2006, Proceedings, 136-144, 2006.01, Mathematical formulas challenge an OCR system with a range of similar-looking characters whose bold, calligraphic, and italic varieties must be recognized distinctly, though the fonts to be used in an article are not known in advance. We describe the use of support vector machines (SVM) to learn and predict about 300 classes of styled characters and symbols..
357. Seiichi Uchida, Akihiro Nomura, Masakazu Suzuki, Quantitative analysis of mathematical documents, International Journal on Document Analysis and Recognition, 10.1007/s10032-005-0142-y, 7, 4, 211-218, 2005.09, Mathematical documents are analyzed from several viewpoints for the development of practical OCR for mathematical and other scientific documents. Specifically, four viewpoints are quantified using a large-scale database of mathematical documents, containing 690,000 manually ground-truthed characters: (i) the number of character categories, (ii) abnormal characters (e.g., touching characters), (iii) character size variation, and (iv) the complexity of the mathematical expressions. The result of these analyses clarifies the difficulties of recognizing mathematical documents and then suggests several promising directions to overcome them..
358. Seiichi Uchida, Masakazu Iwamura, Shinichiro Omachi and Koichi Kise, Data Embedding for Camera-Based Character Recognition, First International Workshop on Camera-Based Document Analysis and Recognition 2005 (CBDAR 2005, Seoul, Korea), 2005.08.
359. Masakazu Iwamura, Seiichi Uchida, Shinichiro Omachi and Koichi Kise, Recognition with Supplementary Information ---How Many Bits Are Lacking for 100% Recognition?---, First International Workshop on Camera-Based Document Analysis and Recognition 2005 (CBDAR 2005, Seoul, Korea), 2005.08.
360. Seiichi Uchida, Hiromitsu Miyazaki, and Hiroaki Sakoe, Mosaicing-by-recognition for recognizing texts captured in multiple video frames, First International Workshop on Camera-Based Document Analysis and Recognition 2005 (CBDAR 2005, Seoul, Korea), 2005.08.
361. Masakazu Suzuki, Seiichi Uchida, and Akihiro Nomura, A ground-truthed mathematical character and symbol image database, Proceedings of 8th International Conference on Document Analysis and Recognition, 10.1109/ICDAR.2005.14, 2005.08.
362. Hiroki Ezaki, Seiichi Uchida, Akira Asano, and Hiroaki Sakoe, Dewarping of document image by global optimization, Proceedings of 8th International Conference on Document Analysis and Recognition, 10.1109/ICDAR.2005.87, 2005.08.
363. Daiki Okumura, Seiichi Uchida, and Hiroaki Sakoe, An HMM implementation for on-line handwriting recognition based on pen-coordinate feature and pen-direction feature, Proceedings of 8th International Conference on Document Analysis and Recognition, 10.1109/ICDAR.2005.50, 2005.08.
364. Hiroto Mitoma, Seiichi Uchida, and Hiroaki Sakoe, Online character recognition based on elastic matching and quadratic discrimination, Proceedings of 8th International Conference on Document Analysis and Recognition, 10.1109/ICDAR.2005.178, 2005.08.
365. Hiromitsu Miyazaki, Seiichi Uchida, and Hiroaki Sakoe, Mosaicing-by-recognition: a technique for video-based text recognition, Proceedings of 8th International Conference on Document Analysis and Recognition, 10.1109/ICDAR.2005.161, 2005.08.
366. Seiichi Uchida and Hiroaki Sakoe, A survey of elastic matching techniques for handwritten character recognition, IEICE Transactions on Information & Systems, 10.1093/ietisy/e88-d.8.1781, 2005.08.
367. Kazuhiko Yamamoto, Shinji Tsuraoka, Hiromichi Fujisawa, Toyohide Watanabe, Hiroshi Murase, Yoshimasa Kimura, Fumitaka Kimura, Masaki Nakagawa, Ryuichi Oka, Norihiro Hagita, Satoshi Naoi, Yasuto Ishitani, Keiji Yamada, Daisuke Nishiwaki, Yoshihiko Hamamoto, Toru Wakahara, Koichi Kise, Shin'ichiro Omachi, Seiichi Uchida, Foreword
Special section on document image understanding and digital documents, IEICE Transactions on Electronics, E88-C, 8, 1779-1780, 2005.08.
368. Seiichi Uchida, Akihiro Nomura, and Masakazu Suzuki, Quantitative analysis of mathematical documents, International Journal on Document Analysis and Recognition, 163-167, vol. 1 of 2, pp. 163-167, 2005.06.
369. Seiichi Uchida and Hiroaki Sakoe, Category-dependent elastic matching based on a linear combination of eigen-deformations, Systems and Computers in Japan, 126-130, 2005.05.
370. Seiichi Uchida, Hiroaki Sakoe, Category-dependent elastic matching based on a linear combination of eigen-deformations, Systems and Computers in Japan, 10.1002/scj.20229, 36, 5, 13-22, 2005.05, A new elastic image matching (EM) technique based on a category-dependent deformation model is proposed. In the deformation model, any deformation of a category is described by a linear combination of eigen-deformations, which are frequent deformation directions of the category and can be estimated statistically from the actual deformations. Experimental results on handwritten characters show that the proposed technique can attain higher recognition rates than conventional EM techniques based on the affine deformation model, which is a typical category-independent deformation model. The results also show the superiority of the proposed technique over those conventional EM techniques in computational efficiency..
371. Akihiro Mori, Seiichi Uchida, Ryo Kurazume, Rin-ichiro Taniguchi, Tsutomo Hasegawa, and Hiroaki Sakoe, Early Recognition of Gestures, 11th Korea-Japan Joint Workshop on Frontiers of Computer Vision, 72-77, 2005.01.
372. Seiichi Uchida, Hiromitsu Miyazaki, Hiroaki Sakoe, Mosaicing-by-recognition for recognizing texts captured in multiple video frames, 1st International Workshop on Camera-Based Document Analysis and Recognition, CBDAR 2005 Proceedings of the 1st International Workshop on Camera-Based Document Analysis and Recognition, CBDAR 2005, 3-9, 2005, Text recognition in video frames is promising because of its following superiorities over text recognition in a still camera image: (1) it is possible to recognize longer texts by concatenating the frames, and (2) it is also possible to improve the quality of the text image by integrating the frames. In this paper, a mosaicing-by-recognition technique is proposed where video mosaicing and text recognition are simultaneously and collaboratively performed in a one-step manner by a dynamic programming-based optimization algorithm. In this optimization algorithm, rotation, scaling, vertical shift, and speed fluctuation of camera motion are efficiently compensated. The results of experiments to evaluate not only the accuracy of text recognition but also that of video mosaicing indicates that the proposed technique is practical and can provide reasonable results in most cases..
373. Hiroto Mitoma, Seiichi Uchida, Hiroaki Sakoe, Online character recognition using eigen-deformations, Proceedings - Ninth International Workshop on Frontiers in Handwriting Recognition, IWFHR-9 2004 Proceedings - Ninth International Workshop on Frontiers in Handwriting Recognition, IWFHR-9 2004, 10.1109/IWFHR.2004.79, 3-8, 2004.12, In online character recognition based on elastic matching, such as dynamic programming matching, many of misrecognitions are often caused by overfitting, which is the phenomenon that the distance between reference pattern of an incorrect category and an input pattern is underestimated by unnatural matching. In this paper, a new recognition technique is proposed where category-specific deformations, called eigen-deformations, are utilized to suppress those misrecognitions. Generally, matching results at overfitting are not consistent with the eigen-deformations. Thus, the overfitting can be detected and penalized by a posterior evaluation of this inconsistency. The result of a recognition experiment showed the usefulness of the proposed technique..
374. Hiroto Mitoma, Seiichi Uchida, and Hiroaki Sakoe, Online character recognition using eigen-deformations, the Ninth International Workshop on Frontiers of Handwriting Recognition, vol. 55, no. 12, pp. 1643-1649, 2004.10.
375. Naoki Matsumoto, Seiichi Uchida, and Hiroaki Sakoe, Prototype setting for elastic matching-based image pattern recognition, Proceedings of 17th IAPR International Conference on Pattern Recognition (ICPR 2004, Cambridge, UK),, 10.1109/ICPR.2004.1334064, 39-43, vol. 1 of 1, pp. 434--438, 2004.08.
376. R. Taniguchi, D. Arita, S. Uchida, R. Kurazume, and T. Hasegawa, Human action sensing for proactive human interface: Computer vision approach, Proceedings of International workshop on Processing Sensory Information for Proactive Systems (PSIPS 2004, Oulu, Finland), Vol.6, No..1,, 2004.06.
377. Eiji Taira, Seiichi Uchida, and Hiroaki Sakoe, Nonuniform slant correction for handwritten word recognition,, IEICE Transactions on Information & Systems, Vol.6, No..1,, 2004.05.
378. Eiji Taira, Seiichi Uchida, and Hiroaki Sakoe, Block boundary detection and title extraction for automatic bookshelf inspection, Tenth Korea-Japan Joint Workshop on Frontiers of Computer Vision (FCV2005, Fukuoka, Japan), Vol.5, No..1,, 2004.02.
379. Eiji Taira, Seiichi Uchida, and Hiroaki Sakoe, A model-based book boundary detection technique for bookshelf image analysis, Asian Conference on Computer Vision (ACCV2004, Jeju Island, Korea), Vol.5, No..1,, 2004.01.
380. Eiji Taira, Seiichi Uchida, Hiroaki Sakoe, Nonuniform slant correction for handwritten word recognition, IEICE Transactions on Information and Systems, E87-D, 5, 1247-1253, 2004.01, Slant correction is a preprocessing technique to improve segmentation and recognition accuracy for handwritten word recognition. All conventional slant correction techniques were performed by the estimation of the average slant angle and the shear transformation. In this paper, a nonuniform slant correction technique for handwritten word recognition is proposed where the slant correction problem is formulated as a global optimal estimation problem of the sequence of local slant angles. The optimal estimation is performed by a dynamic programming based algorithm. From experimental results it was shown that the present technique outperforms conventional uniform slant correction techniques..
381. Masakazu Suzuki, Fumikazu Tamari, Ryoji Fukuda, Seiichi Uchida, Toshihiro Kanahori, INFTY - An integrated OCR system for mathematical documents, Proceedings of the 2003 ACM Symposium on Document Engineering Proceedings of the 2003 ACM Symposium on Document Engineering, 95-104, 2003.12, An integrated OCR system for mathematical documents, called INFTY, is presented. INFTY consists of four procedures, i.e., layout analysis, character recognition, structure analysis of mathematical expressions, and manual error correction. In those procedures, several novel techniques are utilized for better recognition performance. Experimental results on about 500 pages of mathematical documents showed high character recognition rates on both mathematical expressions and ordinary texts, and sufficient performance on the structure analysis of the mathematical expressions..
382. Eiji Taira, Seiichi Uchida, and Hiroaki Sakoe, Book boundary detection from bookshelf image based on model fitting, International Symposium on Information Science and Electrical Engineering, 534-537, 2003.11.
383. Wenjie Cai, Seiichi Uchida, and Hiroaki Sakoe, A comparative study of stroke correspondence search algorithms for online kanji character recognition, International Symposium on Information Science and Electrical Engineering, E83D, 1, 109-111, Vol.1, No.E83, pp.109-111, 2003.11.
384. Seiichi Uchida and Hiroaki Sakoe, A preliminary study of pixel-based motion compensation, International Symposium on Information Science and Electrical Engineering, pp.5, 2003.11.
385. Masakazu Suzuki, Fumikazu Tamari, Ryoji Fukuda, Seiichi Uchida, and Toshihiro Kanahori, INFTY --- An integrated OCR system for mathematical documents, ACM Symposium on Document Engineering (DocEng 2003, Grenoble, France), E82D, 3, 693-700, Vol.3, No.E82, pp.693-700, 2003.11.
386. Seiichi Uchida, Hiroaki Sakoe, Eigen-deformations for elastic matching based handwritten character recognition,, Pattern Recognition, 10.1016/S0031-3203(03)00039-6, Vol.6, No.J81, pp.1251-1258, 2003.09.
387. Seiichi Uchida and Hiroaki Sakoe, Handwritten character recognition using elastic matching based on a class-dependent deformation model, Proceedings of 7th International Conference on Document Analysis and Recognition (ICDAR 2003, Edinburgh, Scotland, 521-524, pp.14, 2003.08.
388. Akihiro Nomura, Kazuyuki Michishita, Seiichi Uchida, and Masakazu Suzuki, Detection and segmentation of touching characters in mathematical expressions, Proceedings of 7th International Conference on Document Analysis and Recognition (ICDAR 2003, Edinburgh, Scotland, Vol.1, No..1, pp.95-100, 2003.08.
389. Akihiro Nomura, Kazuyuki Michishita, Seiichi Uchida, Masakazu Suzuki, Detection and segmentation of touching characters in mathematical expressions, 7th International Conference on Document Analysis and Recognition, ICDAR 2003 Proceedings - 7th International Conference on Document Analysis and Recognition, ICDAR 2003, 10.1109/ICDAR.2003.1227645, 126-130, 2003.01, A technique for the detection and the segmentation of touching characters in mathematical expressions is presented. In the detection stage, a connected component initially recognized into some category is judged as a candidate of touched characters if its feature values deviate from the standard feature values of the category. In the segmentation stage, two component characters of the candidate are decided by the comparison with touching character images synthesized from two single character images. Experimental results showed the effectiveness on the accuracy improvement of the recognition of mathematical expressions..
390. Seiichi Uchida, Hiroaki Sakoe, Handwritten character recognition using elastic matching based on a class-dependent deformation model, 7th International Conference on Document Analysis and Recognition, ICDAR 2003 Proceedings - 7th International Conference on Document Analysis and Recognition, ICDAR 2003, 10.1109/ICDAR.2003.1227652, 163-167, 2003.01, For handwritten character recognition, a new elastic image matching (EM) technique based on a class-dependent deformation model is proposed. In the deformation model, any deformation of a class is described by a linear combination of eigen-deformations, which are intrinsic deformation directions of the class. The eigen-deformations can be estimated statistically from the actual deformations of handwritten characters. Experimental results show that the proposed technique can attain higher recognition rates than conventional EM techniques based on class-independent deformation models. The results also show the superiority of the proposed technique over those conventional EM techniques in computational efficiency..
391. Seiichi Uchida, Hiroaki Sakoe, A handwritten character recognition method based on unconstrained elastic matching and eigen-deformations, 8th International Workshop on Frontiers in Handwriting Recognition, IWFHR 2002 Proceedings - 8th International Workshop on Frontiers in Handwriting Recognition, IWFHR 2002, 10.1109/IWFHR.2002.1030887, 72-77, 2002.12, A fast elastic matching based handwritten character recognition method is investigated. In the method, an unconstrained elastic matching technique, where the matching is optimized locally and individually on each pixel, is utilized together with its a posteriori evaluation based on the eigen-deformations of handwritten characters. Our experimental results show that high recognition rates can be attained by the present method with feasible computations..
392. Seiichi Uchida and Hiroaki Sakoe, A handwritten character recognition method based on unconstrained elastic matching and eigen-deformations, Proceedings of the Eighth International Workshop on Frontiers of Handwriting Recognition (IWFHR-8, Niagara-on-the-Lake, Ontario, Canada),, 2002.08.
393. Seiichi Uchida, Eiji Taira, and Hiroaki Sakoe, Nonuniform slant correction using dynamic programming, Proceedings of 6th International Conference on Document Analysis and Recognition (ICDAR 2001, Seattle, USA),, 2001.09.
394. Mohammad Asad Ronee, Seiichi Uchida, and Hiroaki Sakoe, Handwritten character recognition using piecewise linear two-dimensional warping, Proceedings of 6th International Conference on Document Analysis and Recognition (ICDAR 2001, Seattle, USA),, 2001.09.
395. R. Bogush, S. Maltsev, S. Ablameyko, S. Uchida, and S. Kamata, An efficient correlation computation method for binary images based on matrix factorisation, Proceedings of 6th International Conference on Document Analysis and Recognition (ICDAR 2001, Seattle, USA),, 2001.09.
396. R. Bogush, S. Maltsev, S. Ablameyko, S. Uchida, S. Kamata, An efficient correlation computation method for binary images based on matrix factorisation, 6th International Conference on Document Analysis and Recognition, ICDAR 2001 Proceedings - 6th International Conference on Document Analysis and Recognition, ICDAR 2001, 10.1109/ICDAR.2001.953805, 312-316, 2001.01, A novel algorithm for complexity reduction in binary image processing, namely for computation of correlation between image and object template is proposed. This algorithm is based on direct computation of vector-matrix multiplication with utilisation of binary matrix factorisation approach. Comparison with other algorithms is given and it is shown that our approach allows to reduce time and complexity of this task..
397. Mohammad Asad Ronee, Seiichi Uchida, Hiroaki Sakoe, Handwritten character recognition using piecewise linear two-dimensional warping, 6th International Conference on Document Analysis and Recognition, ICDAR 2001 Proceedings - 6th International Conference on Document Analysis and Recognition, ICDAR 2001, 10.1109/ICDAR.2001.953751, 39-43, 2001.01, In this paper, the effectiveness of piecewise linear two-dimensional warping, a dynamic programming-based elastic image matching technique, in handwritten character recognition is investigated. The present technique is capable of providing compensation for most variations in character patterns with tractable computation. The superiority of the present technique over several conventional two-dimensional warping techniques in variation compensating is experimentally justified. Another comparison with mono-tonic and continuous two-dimensional warping, a more flexible matching technique, reveals that the present method takes far less computation than the latter, yet provides almost the same recognition accuracy for most categories..
398. Seiichi Uchida, Eiji Taira, Hiroaki Sakoe, Nonuniform slant correction using dynamic programming, 6th International Conference on Document Analysis and Recognition, ICDAR 2001 Proceedings - 6th International Conference on Document Analysis and Recognition, ICDAR 2001, 10.1109/ICDAR.2001.953827, 434-438, 2001.01, Slant correction is an indispensable technique for handwritten word recognition systems. Conventional slant correction techniques estimate the average slant angle of component characters and then correct the slant uniformly. Thus these conventional techniques will perform successfully under the assumption that each word is written with a constant slant. However, it is more widely acceptable assumption that the slant angle fluctuates during writing a word. In this paper, a nonuniform slant correction technique is presented where the slant correction problem is formulated as an optimal estimaiton problem of local slant angles at all horizontal positions. The optimal estimation is governed by a criterion function and several constraints for the global and local validity of the local angles. The optimal local slant angles which maximize the criterion satisfying the constraints are searched for efficiently by a dynamic programming based algrithm. Experimental results show the advantageous characteristics of the present technique over the uniform slant correction techniques..
399. Eiji Taira, Toshiyuki Ishida, Seiichi Uchida, Hiroaki Sakoe, Slant correction of handwritten word using two-dimensional warping, Research Reports on Information Science and Electrical Engineering of Kyushu University, 6, 1, 96-97, 2001.01, Slant correction of characters is necessary in the segmentation of a handwritten word into component characters. Conventional slant correction techniques estimate the average slant angle of component characters and correct only uniform slant resulting residual error for each character. In this paper, a slant correction technique which can well correct nonuniform slant is proposed. In the present technique, the slant correction problem is formulated as a non-linear mapping problem of slanted strokes onto vertical straight lines. Then a dynamic programming-based two-dimentional warping algorithm is applied to optimize the mapping. The effectiveness of the present technique was shown by experiments..
400. Kei Yamada, Seiichi Uchida, Hiroaki Sakoe, Speaker normalization based on piecewise linear frequency warping, Research Reports on Information Science and Electrical Engineering of Kyushu University, 6, 1, 91-92, 2001.01, An efficient algorithm for speaker-independent spoken word recognition is presented. This algorithm is based on the time-frequency warping with inter-frame consistency, where each frame of an input pattern is mapped to a reference pattern by controlling the mapping of several points (pivots) on the frame. The mapping of non-pivot points is given by linear interpolation between mapping of two consecutive pivots. The optimal mapping is obtained by using a dynamic programming based algorithm. The computational complexity of the algorithm is reduced to less than that of the previous time-frequency warping algorithm with inter-frame consistency. Experimental results show advantageous characteristics of the present algorithm..
401. Seiichi Uchida, Hiroaki Sakoe, Piecewise linear two-dimensional warping, Proceedings - International Conference on Pattern Recognition, 15, 3, 534-537, 2000.12, A new efficient dynamic programming (DP) algorithm for 2D elastic matching is proposed. The present DP algorithm requires by far less complexity than previous DP-based elastic matching algorithms. This complexity reduction results from piecewise linearization of a 2D-2D mapping which specifies an elastic matching between two given images. Since this linearization can be guided by a priori knowledge related to image patterns to be matched, the present DP algorithm often provides sufficient matching as is shown by experimental results..
402. S. Uchida and H. Sakoe, Piecewise Linear Two-Dimensional Warping, International Conference on Pattern Recognition, 2000.01.
403. S. Uchida and H. Sakoe, An Approximation Algorithm for Two-Dimensional Warping, IEICE Trans. Information & Systems, vol. -D, 2000.01.
404. Seiichi Uchida, Hiroaki Sakoe, An approximation algorithm for two-dimensional warping, IEICE Transactions on Information and Systems, E83-D, 1, 109-111, 2000.01, A new efficient two-dimensional warping algorithm is presented, in which sub-optimal warping is attained by iterating DP-bascd local optimization of warp on partially overlapping subplane sequence. From an experimental comparison with a conventional approximation algorithm based on beam search DP, relative superiority of the proposed algorithm is established..
405. S. Uchida and H. Sakoe, Handwritten Character Recognition Using Monotonic and Continuous Two-Dimensional Warping, Proc. th International Conference on Document Analysis and Recognition, 1999.01.
406. S. Uchida and H. Sakoe, An Efficient Two-Dimensional Warping Algorithm, IEICE Trans. Information & Systems, vol. -D, 1999.01.
407. Seiichi Uchida, Hiroaki Sakoe, An efficient two-dimensional warping algorithm, IEICE Transactions on Information and Systems, E82-D, 3, 693-700, 1999.01, SUMMARY A new dynamic programming (DP) based algorithm for monotonie and continuous two-dimensional warping (2DW) is presented. This algorithm searches for the optimal pixel-to-pixel mapping between a pair of images subject to monotonicity and continuity constraints with by far less time complexity than the algorithm previously reported by the authors. This complexity reduction results from a refinement of the multi-stage decision process representing the 2DW problem. As an implementation technique, a polynomial order approximation algorithm incorporated with beam search is also presented. Theoretical and experimental comparisons show that the present approximation algorithm yields better performance than the previous approximation algorithm..
408. Seiichi Uchida, Hiroaki Sakoe, Handwritten character recognition using monotonic and continuous two-dimensional warping, 5th International Conference on Document Analysis and Recognition, ICDAR 1999 Proceedings of the 5th International Conference on Document Analysis and Recognition, ICDAR 1999, 10.1109/ICDAR.1999.791834, 503-506, 1999.01, In this paper, a handwritten character recognition experiment using a monotonic and continuous two-dimensional warping algorithm is reported. This warping algorithm is based on dynamic programming and searches for the optimal pixel-to-pixel mapping between given two images subject to two-dimensional monotonicity and continuity constraints. Experimental comparisons with rigid matching and local perturbation show the performance superiority of the monotonic and continuous warping in character recognition..
409. S. Uchida and H. Sakoe, A Monotonic and Continuous Two-Dimensional Warping Based on Dynamic Programming, Proc. th International Conference on Pattern Recognition, 1998.01.