researchmap

researchmap

researchmap

researchmap

researchmap

researchmap

受賞論文では，商店街などの公共空間に設置した定点カメラ画像から歩行者間の相対運動を取得することにより，人流の特徴を定量化する手法を提案した．混雑と賑わいを区別することが可能か，という問いに対して，情報科学と産業数学から分野の垣根を超えた研究者が協働することで，主観アンケートや売上といった入手困難な情報を用いずに，賑わいを捉える指標を考するとともに，その妥当性をシミュレーションだけではなく，福岡市，川端通商店街のサポートを受けた実証実験を通して検証している．

researchmap

researchmap

Language：Japanese   Publishing type：Research paper (scientific journal)  

Language：Japanese   Publishing type：Research paper (bulletin of university, research institution)  

Language：Japanese   Publishing type：Research paper (scientific journal)  

Language：English   Publishing type：Research paper (scientific journal)  

Language：English   Publishing type：Research paper (scientific journal)  

Language：Japanese   Publishing type：Research paper (scientific journal)  

Language：Japanese   Publishing type：Research paper (scientific journal)  

Publisher：Multimodal Technologies and Interaction  

Subtitles play a crucial role in facilitating the understanding of visual content when watching films and television programs. In this study, we propose a method for presenting subtitles in a way that considers cognitive load when viewing video content in a non-native language. Subtitles are generally displayed at the bottom of the screen, which causes frequent eye focus switching between subtitles and video, increasing the cognitive load. In our proposed method, we focused on the position, display time, and amount of information contained in the subtitles to reduce the cognitive load and to avoid disturbing the viewer’s concentration. We conducted two experiments to investigate the effects of our proposed subtitle method on gaze distribution, comprehension, and cognitive load during English-language video viewing. Twelve non-native English-speaking subjects participated in the first experiment. The results show that participants’ gazes were more focused around the center of the screen when using our proposed subtitles compared to regular subtitles. Comprehension levels recorded using LightSub were similar, but slightly inferior to those recorded using regular subtitles. However, it was confirmed that most of the participants were viewing the video with a higher cognitive load using the proposed subtitle method. In the second experiment, we investigated subtitles considering connected speech form in English with 18 non-native English speakers. The results revealed that the proposed method, considering connected speech form, demonstrated an improvement in cognitive load during video viewing but it remained higher than that of regular subtitles.

DOI： 10.3390/mti8060051

Web of Science

Scopus

Language：Others   Publishing type：Research paper (scientific journal)   Publisher：IEEE Internet of Things Journal  

Federated learning (FL) has gained widespread attention as a distributed machine learning (ML) technique that offers data protection when training on local devices. Unlike conventional centralized training in traditional ML, FL incorporates privacy and security measures as it does not share raw data between the client and server, thereby safeguarding potentially sensitive information. However, there are still vulnerabilities in the FL field, and commonly used approaches, such as encryption and blockchain technologies, often result in significant computational and communication costs, making them impractical for devices with restricted resources. To tackle this challenge, we present a privacy-preserving FL system specifically designed for resource-constrained devices, leveraging compressive sensing and differential privacy (DP) techniques. We implemented the weight-pruning-based compressive sensing method with an adaptive compression ratio based on resource availability. In addition, we employ DP to introduce noise to the gradient before sending it to a central server for aggregation, thereby protecting the gradient's privacy. Evaluation results demonstrate that our proposed method achieves slightly better accuracy when compared to state-of-the-art methods like DP-federated averaging, DP-FedOpt, and adaptive Gaussian clipping-DP (AGC-DP) for the MNIST, Fashion-MNIST, and Human Activity Recognition data sets. Furthermore, our approach achieves this higher accuracy with a lower total communication cost and training time than the current state-of-the-art methods. Moreover, we comprehensively evaluate our method's resilience against poisoning attacks, revealing its better resistance than existing state-of-the-art approaches.

DOI： 10.1109/JIOT.2023.3347552

Web of Science

Scopus

researchmap

Language：Others   Publishing type：Research paper (scientific journal)   Publisher：IEEE Access  

In modern transportation, autonomous driving technology has emerged as a significant innovation, offering improved safety, efficiency, and convenience. One of its core technologies is path planning, which creates an optimal route for vehicles in various traffic situations, ensuring they arrive safely and efficiently. However, existing path planning methods suffer from inflexible and inefficient road sampling, and limitations in path expression due to curve models. In this paper, we introduce a novel path planning framework that operates from a coarse-to-optimal approach. We first introduce an adaptive sampling method to construct a three-dimensional (3D) space-aware profiling map in the Frenet frame to quantify abstract traffic scenarios. This cost map guides the generation of waypoints by incorporating obstacle distributions in the s -direction and cost values in the l -direction. We then design a cost function to obtain the coarse path by considering ride comfort, anti-deviation from the center of the road, and path safety. Finally, a new path model is re-established based on the Taylor series, and a quadratic programming approach is employed to fine-tune the coarse path to attain the optimal path. Experimental results show that the proposed adaptive sampling method can dynamically adjust the sampling strategy according to the change in the number and distribution of obstacles on the road, significantly improving the sampling effectiveness rate to 100%, 99.85%, 99.68% and 99.61% for one-obstacle, two-obstacle, three-obstacle, and four-obstacle scenarios. Compared with the conventional method, the generated path length is shortened by 3.48% and 8.42%, and the maximum heading angle is reduced by 18.68% and 30.21% in the scenarios with obstacles of the same and different sizes.

DOI： 10.1109/ACCESS.2023.3347399

Web of Science

Scopus

researchmap

Publisher：2024 IEEE International Conference on Pervasive Computing and Communications Workshops and other Affiliated Events, PerCom Workshops 2024  

In recent years, with the trend towards shorter working hours, the quality of meetings has increasingly impacted these hours. Consequently, meeting quality and value become more important. To achieve better results in limited time, everyone needs to conduct meetings effectively. This requires skills such as facilitating the meeting, listening to others' opinions, accurately expressing one's own views, and using body language. However, these skills are currently only correlated with meeting effectiveness and lack comprehensive qualitative and quantitative assessments, leaving specific methods for improvement unclear. Additionally, the type of meeting can lead to participant stress or disinterest, making it essential to understand their psychological safety and engagement for more effective meetings. Measuring these psychological states poses significant challenges due to privacy and compliance concerns, particularly when using cameras or wearable sensors. This paper broadly addresses these issues and reports on our approach to detecting non-verbal cues, specifically nodding, from chair movements as a potential solution.

DOI： 10.1109/PerComWorkshops59983.2024.10502817

Web of Science

Scopus

Publisher：MOBISYS 2024 - Proceedings of the 2024 22nd Annual International Conference on Mobile Systems, Applications and Services  

Since office work has become large-scale and diversified in companies or organizations, work engagement and efficiency have been always an important index of a team's or group's evaluation because it is directly connected to their outcomes. In order to identify the group work context, we first need to recognize for what and how long the individual members are spending their time at their desks, but without privacy concerns and underestimation of their actual work. In this paper, we propose and evaluate the base system of personal desk activity recognition by using a low-cost compact WiFi node and its WiFi channel state information (CSI), which can lead to a lightweight group work context identification system. As a result, we achieved 94.2% desk activity recognition accuracy using the on-desk receiver, in recognizing five different classes.

DOI： 10.1145/3643832.3661452

Web of Science

Scopus

Publisher：MOBISYS 2024 - Proceedings of the 2024 22nd Annual International Conference on Mobile Systems, Applications and Services  

Indoor activity recognition using WiFi sensing is expected to have a wide range of applications, such as monitoring the elderly and home security. The state of radio wave propagation is called Channel State Information (CSI) and can be obtained using specific devices. By collecting CSI and applying machine learning, it is possible to recognize activities. However, CSI is sensitive to changes in the environment, so whenever the arrangement of furniture or the layout of the room changes, it is necessary to re-collect sample data and retrain the model. Retraining a model requires annotation work, which is costly in terms of time and effort. To address this issue, this paper proposes an annotation system that uses backscatter tags to reduce the cost of data collection and model training. In this system, a backscatter tag that generates a frequency shift depending on its angle is attached to a person during data collection, and activity recognition is performed by detecting the presence of the frequency shift. The backscatter tag-based recognition results are then used as pseudo-ground truth for model update.

DOI： 10.1145/3643832.3661451

Web of Science

Scopus

Publisher：IEEE Access  

Machine learning's exponential growth has transformed healthcare, with Federated Learning (FL) playing a pivotal role. Despite its significance, FL is vulnerable to privacy attacks. In response, researchers have integrated differential privacy (DP) into FL. Nevertheless, incorporating DP introduces challenges such as increased total communication costs and computational overheads due to the introduction of noise. This drawback renders FL with DP less viable for healthcare systems, characterized by numerous low-resource devices and network bandwidth constraints. To overcome this limitation, we propose integrating a Discrete Fourier Transform (DFT) aggregator post-noise addition to transform the gradient generated by local training before sending it to the central server. This process reduces the gradient size and provides rudimentary encryption. The evaluation results reveal the superior performance of our proposed method, demonstrating an enhanced accuracy ranging from 0.2% to 2% compared to existing differential privacy techniques, including RDP, DP-SGD, ZcDP, LDP-Fed, and DP-AdapClip. Our approach substantially reduces the total communication costs (ranging from 6% to 43% across different privacy budgets) with faster training times in healthcare datasets such as the PIMA Indian database and Breast Cancer Histopathology Images.

DOI： 10.1109/ACCESS.2024.3418016

Web of Science

Scopus

Publisher：International Conference on Information Networking  

As the high-speed wireless network spreads, the usage of shared Wi-Fi in places such as offices, coworking spaces, and homes has been increased. Along with the development and diversification of digital content, the amount of data consumed by individual devices has also grown. In situations where multiple users share limited network resources, some content that consumes a large amount of bandwidth like video streaming can potentially degrade the Quality of Experience (QoE) for other users nearby. Consequently, a need for persuasive intervention systems that encourage prosocial behavior is rising taking into consideration the QoE of other network users. To address this issue, this study proposes a "Persuasive Public Wi-Fi"that employs Wi-Fi to incrementally convince users towards more considerate usage of network resources. To be specific, we suppose that the persuasive public Wi-Fi must include three different modes: 1) a mode of normal networking, 2) a mode that can intervene with individual users through a captive portal, and 3) a mode that intentionally limits bandwidth, i.e., Quality of Service (QoS) control. This work demonstrates the design and implementation of the proposed system and presents the results of operational verification using a prototype system.

DOI： 10.1109/ICOIN59985.2024.10572181

Scopus

Publisher：MOBISYS 2024 - Proceedings of the 2024 22nd Annual International Conference on Mobile Systems, Applications and Services  

We present a secure decentralized learning framework suitable for resource-constrained devices within a cluster environment. Our approach focuses on enhancing privacy preservation during model aggregation by utilizing Differential Privacy. This technique adds random noise to gradients obtained from local training on edge devices before sending them for aggregation. This noise addition ensures that sensitive information within the gradients remains distorted, thus safeguarding user privacy. We showcase the implementation of our system on a cluster system employing Raspberry Pi 4 Model B devices, illustrating its feasibility and effectiveness in real-world scenarios. Through this demonstration, we highlight the practical applicability of our system in enabling secure decentralized learning within resource-constrained environments.

DOI： 10.1145/3643832.3661843

Web of Science

Scopus

Publisher：CEUR Workshop Proceedings  

In an era where digital technology profoundly influences our daily habits, we explore the potential of color psychology as a key element in the design of persuasive technology to promote health behavior change. This paper presents a strategy that uses color psychology to both encourage positive behaviors and discourage negative ones. We present two different applications: "eat2pic," a system designed to promote healthier eating habits by transforming meal selection into an engaging activity of adding color to images, and "color-wall," an application designed to reduce digital distractions by applying a grayscale filter to non-essential digital content. Through these demonstrations, we explore the efficacy and potential of color psychology in persuasive technology to influence user behavior in a health context and offer promising solutions to improve dietary choices and concentration while using computers.

Scopus

Language：English   Publishing type：Research paper (scientific journal)   Publisher：IEEE Access  

The conventional positioning methods, such as fingerprinting and trilateration, are commonly used in existing Wi-Fi positioning systems. Although the fingerprinting method offers relatively high accuracy, it faces challenges due to its sensitivity to environmental changes and the necessity of extensive training data and calibration. The trilateration method calculates positions based on the distances between anchors and targets. However, inaccuracies in measuring these distances could significantly impact the overall accuracy. Additionally, the necessity for at least three anchors creates a requirement for a more extensive infrastructure, posing challenges to practical service deployment. In this paper, we introduce <italic>bilatangulation</italic>, a novel cluster-based double-step positioning method that leverages distances calculated using fine timing measurement (FTM) and angles determined using channel state information (CSI) from two anchors. The first step addresses the symmetry problem of the two intersections in distance-based positioning by utilizing the angle orientations. In the second step, we performed measurement error compensation by clustering multiple intersections generated from both distance and angle data, taking into account the characteristics of each cluster. Our practical experiment was conducted indoors using off-the-shelf network interface card (NIC). For positioning, only two anchors were employed, resulting in an original mean positioning error (MPE) of 1.58 <italic>m</italic>. Applying a measurement error compensation step reduced the final MPE by 88% compared to the original MPE.

DOI： 10.1109/ACCESS.2024.3447112

Web of Science

Scopus

Language：English   Publisher：International Institute of Applied Informatics  

DOI： 10.52731/liir.v005.302

CiNii Research

Publisher：2024 8th International Conference on Robotics, Control and Automation, ICRCA 2024  

Automated driving technology has garnered increasing attention due to its potential to reduce traffic accidents, alleviate congestion, and enhance travel convenience. As a key component of automated driving technology, trajectory planning plays a pivotal role in determining feasible trajectories in different traffic scenarios. However, there is currently insufficient research on the performance of different curve models in lane-changing scenarios. Furthermore, the conventional speed planning method (DP&QP) often ignores vehicle kinematics and traffic scenarios when generating convex spaces, resulting in inefficiency and longer run time. In path planning, this paper compared the performance of Dubins, Sine, Bézier, and B-spline curves in lane change scenarios, ultimately selecting the B-spline curve as the lane change path model to ensure that the vehicle remains close to the center line of the road after a lane change. In speed planning, we introduce a novel method for directly generating a convex space, taking into account the kinematics of the ego vehicle and the current traffic conditions. This approach allows for adaptive adjustments to the speed planning scheme based on the traffic situation. Compared with the conventional trajectory planning method, this approach significantly reduces the run time by 32.7%.

DOI： 10.1109/ICRCA60878.2024.10649189

Scopus

Language：English   Publisher：International Institute of Applied Informatics  

DOI： 10.52731/liir.v005.312

CiNii Research

Publisher：2024 8th International Conference on Robotics, Control and Automation, ICRCA 2024  

Dense video captioning tasks are used to detect interesting events and provide descriptive text for these events from untrimmed videos. This technology has the potential to be used in security surveillance and human care applications. However, current methods often overlook the relationships between objects in the video, which limits their applicability and makes it challenging to adapt them to specific domains, such as video summarization for indoor human activities. In these scenarios, human activities are closely intertwined with the objects in the scene. In this paper, we propose a plug-and-play module designed to enhance existing dense video captioning methods with spatial information. Specifically, we extract spatial information about the interesting objects using Red-Green-Blue-Depth (RGB-D) images and the results of image segmentation. We then integrate this information into the captions generated by the Dense Video Captioning (DVC) method using a fine-tuned Large Language Model (LLM). We evaluate the performance of our model on a custom dataset and demonstrate that our system provides a convenient and effective approach for obtaining space-enhanced captions.

DOI： 10.1109/ICRCA60878.2024.10649311

Scopus

Language：Others   Publishing type：Research paper (other academic)   Publisher：Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)  

Understanding user emotion is essential for Human-AI Interaction (HAI). Thus far, many approaches have been studied to recognize emotion from signals of various physiological modalities such as cardiac activity and skin conductance. However, little attention has been paid to the fact that physiological signals are influenced by and reflect various factors that have little or no association with emotion. While emotion is a cross-modal factor that triggers responses across multiple physiological modalities, features used in existing approaches also reflect modality-specific factors that affect only a single modality and have little association with emotion. To address this, we propose an approach to extract features that exclusively reflect cross-modal factors from multimodal physiological signals. Our approach introduces a multilayer RNN with two types of layers: multiple Modality-Specific Layers (MSLs) for modeling physiological activity in individual modalities and a single Cross-Modal Layer (CML) for modeling the process by which emotion affects physiological activity. By having all MSLs update their hidden states using the CML hidden states, our RNN causes the CML to learn cross-modal factors. Using real physiological signals, we confirmed that the features extracted by our RNN reflected emotions to a significantly greater extent than the features of existing approaches.

DOI： 10.1007/978-981-99-7019-3_40

Scopus

researchmap

Other Link： https://dblp.uni-trier.de/db/conf/pricai/pricai2023-1.html#IshikawaKNNIMA22

Language：Others   Publishing type：Research paper (other academic)   Publisher：ACM International Conference Proceeding Series  

In this demo, we will show our proposed system that can count nodding without either a camera or any sensor attached to the person. Our proposed system capitalizes on the fact that the upper body moves in conjunction with nodding and that this body motion slightly shakes the chair. We explore the challenge of recognizing nodding from the extremely subtle sway of a chair. To recognize nods in real-Time, we employed a supervised learning approach using acceleration data from sensors attached to the chair's backrest. Ultimately, the Support Vector Machine (SVM) achieved a nodding recognition accuracy of 0.990. Further testing of the accuracy of nodding frequency measurements yielded an accuracy of 0.947, suggesting that the optimal position for the accelerometer is the backrest. These results indicate that simply placing the accelerometer on the backrest can effectively quantify the frequency of nods from seated participants.

DOI： 10.1145/3627050.3630740

Scopus

researchmap

Language：Others   Publishing type：Research paper (other academic)   Publisher：2023 14th International Conference on Mobile Computing and Ubiquitous Network, ICMU 2023  

Aiming to evaluate the concentration and engagement level in offline group discussion scenarios, we seek to capture the interaction information among meeting participants. In this paper, we focus on the facial orientation and present an end-To-end hybrid system capable of 2D overhead facial orientation estimation. Compared to common perspectives from which personal identity information is exposed, we explore approaches to perform overhead facial orientation estimation due to the privacy problem. The proposed system is a neural network-based architecture and composed of three modules. First we leverage and train from scratch the YOLOv8 model on the HollywoodHeads dataset, which allows full-perspective head detection in visual scenes. With the obtained head location, we then generate the corresponding head masks from the visual information within the bounding box. Finally, we fit them to ellipse so that 2D overhead facial orientation of all the meeting participants in the visual scenes can be estimated with the ellipse properties. Experiments demonstrate that our system achieves an MAE of 12.47 on the test dataset.

DOI： 10.23919/ICMU58504.2023.10412258

Scopus

researchmap

Language：Others   Publishing type：Research paper (other academic)   Publisher：2023 14th International Conference on Mobile Computing and Ubiquitous Network, ICMU 2023  

In this paper, we propose a method to measure human head motion, especially nodding, without attaching any sensors to the person. Our proposed system focuses on the fact that the upper body moves along with nodding and that the body motion slightly shakes the chair. We challenge the problem of whether it is possible to recognize a nodding from the extremely slight sway of a chair. To reveal the optimal position of sensors, we collected data by attaching multiple accelerometers to various positions on a chair, including the backrest, the seat's underside, and the legs. Using a supervised learning approach, we determined the best positions and combinations of sensors for recognizing nodding more collectively. The Support Vector Machine (SVM) achieved a nodding recognition accuracy of 0.990. Further testing of the accuracy of nodding frequency measurements resulted in an accuracy of 0.947, suggesting that the best position for the accelerometer is the backrest. These results suggest that simply placing the accelerometer on the backrest can effectively quantify the nod frequency of seated participants.

DOI： 10.23919/ICMU58504.2023.10412249

Scopus

researchmap

Language：Others   Publishing type：Research paper (scientific journal)   Publisher：IEEE Access  

Quadriplegics face a communication obstacle as their physical abilities are restricted, leaving them unable to speak or use their limbs, with only their upper neck being mobile. So, we propose a recognition system and a new communication language utilizing Morse code and head movements, to break this barrier. We aim to overcome the limitations of camera-based and wearable-sensor methods, including occlusion, privacy concerns, and user inconvenience. The goal is to passively detect quadriplegics' head movements and map them to their corresponding character. The dataset including all 26 alphabet letters, was gathered in various settings, including single-user and multi-human environments, with multiple locations for each setting. For evaluation, 2% samples are randomly selected from the unseen environment to be used with the seen environment as a training dataset. Based on the results, our system demonstrates practical feasibility for real-world implementation, with accuracy rates of 94% and 80% achieved in single-user and multi-human environments, respectively.

DOI： 10.1109/ACCESS.2023.3326259

Web of Science

Scopus

researchmap

Language：Others   Publishing type：Research paper (other academic)   Publisher：2023 14th International Conference on Mobile Computing and Ubiquitous Network, ICMU 2023  

In recent years, many studies explore CSI-based Wi-Fi sensing produce high accurate results in device-free localization, human activity recognition. Channel state information (CSI) represents how wireless signals propagate from transmitter to receiver and provides rich information to identify human presence. By collecting CSI from the specific devices, the machine learning model can be trained to recognize the human activity. However, in multi-room residential settings where walls and furniture obstruct signals, effective coverage with a limited number of transceivers becomes a crucial challenge, underlining the importance of their optimal placement. In this paper, we deployed multiple transceivers in a smart home environment in our university and studied transceiver arrangement patterns for CSI-based indoor localization and activity recognition. We compared the accuracy of localization and activity recognition for a total of 18 patterns consisting of up to five transceivers. In the activity recognition, we used Support Vector Machine (SVM) to classify whether or not a person is moving. The results show that the pattern using only one pair of transceivers achieved an accuracy 85% and covered the entire house. Meanwhile, in the localization we used Light Gradient Boosting Machine (LGBM) to classify which room the person is. The results show that accu-racy decreases as the number of devices is reduced. Therefore, we investigated deployment pattern to achieve accuracy with smaller number of transceivers.

DOI： 10.23919/ICMU58504.2023.10412230

Scopus

researchmap

Language：Others   Publishing type：Research paper (other academic)   Publisher：Proceedings - International Conference on Tools with Artificial Intelligence, ICTAI  

In this paper, we propose ToonMeet, a hybrid frame-work for high-resolution and style-controllable online meeting toonification that ensures real-time operation speed. ToonMeet applies video frame interpolation to traditional portrait toonification pipelines, allowing for the synthesis of intermediate frames between adjacent toonified keyframes, significantly accelerating the overall process and saving computational resources. However, this approach brings a new problem, where prevailing flow-based video frame interpolation methods tend to cause more ghost and blur artifacts in toonified scenes compared to non-toonified scenes, especially when fast-moving objects exist. We study this previously undiscussed problem and explore its causes. To address this, we introduce a new dataset called TM3B (Toonified Multi-modal Meeting Behaviors), offering high-resolution and cross-platform multi-modal stylized meeting data of Japanese youth in various scenarios. Then, we fine-tune ToonMeet on these tailored data and the resulting model presents improved optical flow estimation ability on toonified videos. Extensive experiments demonstrate that ToonMeet can achieve great spatiotemporal performance and perform high-quality toonification of online meetings with real-time operation speed.

DOI： 10.1109/ICTAI59109.2023.00013

Web of Science

Scopus

researchmap

Language：Others   Publishing type：Research paper (other academic)   Publisher：2023 14th International Conference on Mobile Computing and Ubiquitous Network, ICMU 2023  

Nowadays, there is a growing demand to understand the mental well-being of office workers, driven by increased awareness of its impact on productivity and the need for healthier work environments. Recently, the use of Wi-Fi channel state information (CSI) for activity recognition has received significant attention due to its wide availability and privacy protection. In this paper, we propose a passive desk body gesture recognition system that utilizes Wi-Fi CSI from an ESP32 toolkit to automatically detect the worker's mood and emotions. The system is designed to operate within the Internet of Things ecosystem, employing a low-energy device to collect and compress CSI measurements, resulting in improved energy efficiency and cost-effectiveness. The proposed system demonstrates high recognition accuracy of over 98 % in-session and 72 % out-session evaluations.

DOI： 10.23919/ICMU58504.2023.10412222

Scopus

researchmap

Language：Others   Publishing type：Research paper (other academic)   Publisher：2023 14th International Conference on Mobile Computing and Ubiquitous Network, ICMU 2023  

Nowadays more and more people are wearing smart-watches in their daily lives. The various sensors embedded in smartwatches bring the ability to evaluate users' status as well as the risk of privacy issues. For example, if users are typing on key-boards while wearing smartwatches, the attacker could know the typing contents from the sensor data collected by the malicious applications that are installed on the targets' smartwatches. In this paper, we propose WatchLogger, the framework using audio and accelerometer signals to recognize the English words being typed, for demonstrating how to implement the smartwatch-based side-channel attack. Different from the previous studies that focused on the recognition of each key or pair of keys being pressed, WatchLogger aims to perform recognition on the scale of words. To achieve this goal, WatchLogger exploits the audio signals for segmentation and the accelerometer signals for classification. In addition, we propose an ensemble classification model to deal with the problem caused by too many words. At last, we build the dataset WTW-100 with 100 classes of words and 100 samples for each class, and we conduct experiments on the dataset. The experimental results show an accuracy of 98.5 % for keystroke recognition and 91.5 % for word classification, showing a considerable performance of WatchLogger.

DOI： 10.23919/ICMU58504.2023.10412218

Scopus

researchmap

Language：Others   Publishing type：Research paper (other academic)   Publisher：2023 14th International Conference on Mobile Computing and Ubiquitous Network, ICMU 2023  

Internet use is expanding along with the spread of smartphones and the promotion of digital use, but information security literacy is required to use the Internet safely. This paper addresses these challenges by focusing on the potential of microlearning and nudge theory to support continuous learning to solve these problems and presents a novel learning system that utilizes a Wi-Fi landing portal that appears as a dialog or a browser. To assess the effectiveness of the proposed system, we conducted a comparative analysis of information security test scores before and after the operation of the learning system and questionnaire-based evaluation. The results showed that the analysis of answer scores before and after the operation of the learning system was significant (\mathrm{p} < 0.05), indicating an improvement in scores. It was also found that using interface information with nudges effectively motivated participants to engage in repeated learning behaviors. However, it should be noted that frequent interventions might reduce acceptability, and the psychological burden arising from task interruptions needs to be taken into account.

DOI： 10.23919/ICMU58504.2023.10412210

Scopus

researchmap

Language：Others   Publishing type：Research paper (other academic)   Publisher：Proceedings - 2023 22nd IEEE/WIC International Conference on Web Intelligence and Intelligent Agent Technology, WI-IAT 2023  

When a tourist wants to visit a site, he/she first asks about the category of the site. The availability of detailed information about the site allows him/her to tailor his/her visit to his/her preferences. This information is therefore essential for Point of Interest (POI) recommendation. However, it is rarely or never available for lesser-known POIs. Lesser-known POIs can be considered as places or events that local people know well, and that may be important to them, but that others do not know about. We propose an approach to estimate the categories of lesser-known POIs based on information from social media. The originality of this approach lies in the extraction of information and links between them, the encoding of the POIs, the representation of the data, and the combination of machine learning techniques such as Few Shot Learning, LightGCN, and Clustering for the estimation of POI categories. The results of the experiments would allow us to confirm that our approach can estimate POI categories and thus discover information about POIs that may be relevant. This approach would be useful for our future work on POI recommendations.

DOI： 10.1109/WI-IAT59888.2023.00020

Web of Science

Scopus

researchmap

Language：Others   Publishing type：Research paper (scientific journal)   Publisher：IEEE Access  

Autonomous vehicles are an essential tool for promoting the development of intelligent transportation systems (ITS) and can effectively reduce traffic accidents caused by human errors. As an important part of the automatic driving software system, path planning is responsible for generating the motion trajectory of the vehicle, which is the primary factor determining driving quality. However, solution space construction and optimization problem formulation remain challenging research areas in the field of path planning. In this paper, we propose a multi-objective optimization algorithm for static obstacle avoidance to improve the comfort, safety and anti-deviation of the planned trajectory. We decouple the lateral and longitudinal motion of the vehicle using the Frenet frame and discretize the driving state space to generate target states of the vehicle. Based on the initial and target states, we generate a set of lateral and longitudinal motion trajectories using quintic and quartic polynomials, respectively. In addition, we design a cost function that comprehensively considers the comfort, safety, and deviation distance of the road center line by combining an acceleration check, curvature check, and collision check. As part of the cost function, we propose a novel method to quantify the safety of candidate trajectories considering the size of obstacles. The experimental results show that the proposed algorithm can quantize the safety of candidate paths and improve comfort 13.47%, 32.19%, 59.36% and 18.60% on a straight road, curvy road, intersection and U-shaped road, respectively. Furthermore, the algorithm can improve anti-deviation by 63.72%, 13.86%, 44.36%, and 45.56% on a straight road, curvy road, intersection and U-shaped road, respectively.

DOI： 10.1109/ACCESS.2023.3294713

Web of Science

Scopus

researchmap

Language：Others   Publishing type：Research paper (other academic)   Publisher：Proceedings - IEEE International Conference on Mobile Data Management  

Federated learning provides techniques for training algorithms using mobile or decentralized devices, in contrast to traditional machine learning in which algorithm training is performed on centralized devices. In addition, federated learning provides privacy and security features, as the client and server do not share raw data, which may contain confidential information. A number of studies have shown, however, that using federated learning alone is not enough to protect data privacy in certain situations. To overcome this problem, differential privacy is proposed, which is a technique in which artificial noise is added to the raw data. By implementing this method, a high level of privacy protection can be obtained, however this added noise also reduces model accuracy. To address this issue, this paper proposes a new approach to implement differential privacy in federated learning using adaptive Gaussian clipping. We implemented the method by tightening the privacy budget, and introducing dynamic sampling probability, adaptive clipping based on hyperparameters, and a new privacy loss calculation. Our method's main objective is to adaptively change the amount of noise given to the model, thereby maximizing the model's accuracy performance, while maintaining privacy protection levels. Evaluation results show that our proposed method presents slightly better accuracy when compared to other existing differential privacy variants such as RDP, DP-SGD, and ZcDP, for both balanced (i.i.d.) and unbalanced datasets (non-i.i.d.), for a lower total communication cost than some variants.

DOI： 10.1109/MDM58254.2023.00042

Scopus

researchmap

Other Link： https://dblp.uni-trier.de/db/conf/mdm/mdm2023.html#HidayatNDA23

Language：Others   Publishing type：Research paper (other academic)   Publisher：Proceedings - IEEE International Conference on Mobile Data Management  

Federated learning has gained popularity as a distributed machine learning approach that provides security and privacy for data trained on local devices. However, vulnerabilities still exist in this approach, and common solutions such as encryption and blockchain techniques often suffer from high computation and communication costs, making them impractical for resource-constrained devices. To solve this problem, we propose a privacy-preserving federated learning system that leverages compressive sensing and differential privacy, specifically designed for devices with limited computational resources. In this paper, we demonstrate the capabilities of our proposed system in resource-limited environments. We outline the features, infrastructure, and algorithm of our proposed system, and simulate its performance using image datasets on a Raspberry Pi 4 and an Android smartphone in a cloud environment. Our approach offers a practical solution for secure and privacy-preserving federated learning in resource-constrained scenarios, with potential applications in various domains such as healthcare, IoT, and edge computing.

DOI： 10.1109/MDM58254.2023.00038

Scopus

researchmap

Other Link： https://dblp.uni-trier.de/db/conf/mdm/mdm2023.html#HidayatNA23

Language：Others   Publishing type：Research paper (other academic)   Publisher：2023 International Conference on Consumer Electronics - Taiwan, ICCE-Taiwan 2023 - Proceedings  

Human activity recognition (HAR) using the deep learning method has caught the attention of researchers thanks to its automatic feature extraction and accurate prediction capabilities. However, for applications based on a wearable sensor, such as an inertial measurement unit (IMU), the process of collecting and hand-labeling large amounts of data is complicated and labor-intensive, meaning that there is a limited amount of data available for model training. Therefore, there is a need to propose and develop data augmentation approaches to generate high quality data for the growth of HAR research. We propose a head-mounted virtual gyroscope signal generator to alleviate the problems caused by the lack of data in head movement-related applications. Unlike previous work, our system only generates head-motion related gyroscope data, minimizing system complexity. We trained a deep-learning model in a head motion-based application with different generated sensor data ratios, and show the viability of our proposed data generation method.

DOI： 10.1109/ICCE-Taiwan58799.2023.10227010

Scopus

researchmap

Other Link： https://dblp.uni-trier.de/db/conf/icce-tw/icce-tw2023.html#LuCDNA23

Language：Others   Publishing type：Research paper (other academic)   Publisher：Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)  

In this paper, we propose a purchasing support system for offline stores that can present information to users without being evasive by interactively presenting a small amount of product information according to their interest in the product as predicted by the system. We design and implement a proof-of-concept system for an actual use case scenario: a snack purchasing area in an office. We evaluated the effectiveness of the proposed system through a snack purchasing experiment with 11 participants. Experimental results suggest that the proposed method can induce user interest and encourage viewing and that some users may receive and consider the information presented while making a purchase.

DOI： 10.1007/978-3-031-30933-5_8

Scopus

researchmap

Language：Others   Publishing type：Research paper (other academic)  

In this study, the target behavior of prosocial behavior was defined as garbage disposal (behavior of taking out the trash), and an intervention method using an inducement was proposed and evaluated. Specifically, we conducted an intervention in which the saturation of a trash can was visualized using a colored light that could be changed to red, yellow, and green. In addition, questionnaire surveys were administered before and after the intervention to 28 undergraduate and graduate students and faculty members, asking about their attitudes and feelings toward garbage disposal behavior, and personality traits. The results of a survey on the frequency of trash disposal be-fore and after the experiment showed that the time that the saturated trash can was left unattended was reduced by 81&#37;. The questionnaire survey confirmed that the intervention increased the positive attitude toward trash disposal behavior, indicating that the intervention had an impact on trash disposal behavior. In addition, we were able to confirm the characteristics of individual personality traits and susceptibility to intervention.

Publishing type：Research paper (international conference proceedings)  

In this study, the target behavior of prosocial behavior was defined as garbage disposal (behavior of taking out the trash), and an intervention method using an inducement was proposed and evaluated. Specifically, we conducted an intervention in which the saturation of a trash can was visualized using a colored light that could be changed to red, yellow, and green. In addition, questionnaire surveys were administered before and after the intervention to 28 undergraduate and graduate students and faculty members, asking about their attitudes and feelings toward garbage disposal behavior, and personality traits. The results of a survey on the frequency of trash disposal be-fore and after the experiment showed that the time that the saturated trash can was left unattended was reduced by 81%. The questionnaire survey confirmed that the intervention increased the positive attitude toward trash disposal behavior, indicating that the intervention had an impact on trash disposal behavior. In addition, we were able to confirm the characteristics of individual personality traits and susceptibility to intervention.

Scopus

researchmap

Language：Japanese  

Research on behavioral change by an inducement using an inducement in trash dumping behavior

Language：Others   Publishing type：Research paper (other academic)   Publisher：IEEE Wireless Communications and Networking Conference, WCNC  

Wi-Nod: Head Nodding Recognition by Wi-Fi CSI Toward Communicative Support for Quadriplegics
Recently, the studies of wireless device-free human sensing technology have dramatically advanced with enabling a variety of applications, from activity recognition to vital sign monitoring. In this paper, we propose Wi-Nod which leverages the Wi-Fi Channel State Information (CSI) to detect head nodding gestures for each Morse code symbol based on time-frequency features for accurate recognition accuracy in multi-human context environment. The system consists of three basic modules: data collection, data preprocessing, and learning part based on the inception model. The model was trained to perform the head movement detection based on the CSI spectrogram collected by the ESP32 nodes. We evaluated the performance of the system on four different data sets collected in two different sessions. Our system achieves over 95&#37; recognition accuracy that reveals the feasibility of Wi-Nod system for real-life deployment.

DOI： 10.1109/WCNC55385.2023.10118666

Web of Science

Scopus

researchmap

Language：English   Publishing type：Research paper (other academic)   Publisher：2023 IEEE International Conference on Pervasive Computing and Communications Workshops and other Affiliated Events, PerCom Workshops 2023  

Differential Privacy with Weighted ϵ for Privacy-Preservation in Human Activity Recognition
Many services based on human activity recognition (HAR) have been developed; however, user activity data include a large amount of private information. Although privacy protection is important in activity recognition, it has not been sufficiently explored. Therefore, we propose a privacy-preserving mechanism for HAR services that uses differential privacy. The proposed method reduces the user recognition accuracy to a level that satisfies the privacy requirements by adding weighted noise to the features in the learning model construction and then improves the activity recognition accuracy (service usefulness). The results indicate that when the privacy requirement is defined as less than the probability of a user being identified by chance, the proposed method improves the activity recognition accuracy by approximately 10 &#37; compared to the conventional method.

DOI： 10.1109/PerComWorkshops56833.2023.10150239

Scopus

researchmap

Language：English   Publishing type：Research paper (other academic)  

C-HAR: Compressive Measurement-Based Human Activity Recognition

Authorship：Last author   Language：English   Publishing type：Research paper (international conference proceedings)  

researchmap

Language：Japanese   Publisher：The Japanese Society for Artificial Intelligence  

<p>In recent years, with the spread of wearable devices and their built-in sensors, the analysis of time-series heartbeat data has been promoted in research in the healthcare field. Because time-series heartbeat data can be used to detect abnormal heartbeats and estimate psychological stress, there are many medical and healthcare services that use heartbeat data. However, time-series heartbeat data contains a lot of privacy information, which requires careful consideration of privacy protection for users of these services. In this study, we proposed and evaluated a privacy-preservation mechanism using ``differential privacy'' for a service that detects abnormal heartbeats based on time-series heartbeat data. The evaluation results show that the proposed method is more effective than conventional differential privacy-based mechanisms in balancing service usefulness while realizing user privacy protection.</p>

DOI： 10.11517/pjsai.jsai2023.0_1l4os18a03

CiNii Research

Web of Science

Web of Science

Web of Science

Publisher：ISKE 2023 - 18th International Conference on Intelligent Systems and Knowledge Engineering  

Local events are important for the vitality and dynamism of a city. However, some events or sites are lesser-known or only known by the local community. The information about them is then limited or not available. To address this issue, we propose an approach to automatically generate knowledge graphs from information posted on social networks and data on human mobility. This graph is generated from the tweets using the syntactic relationships between the verbs, nouns, and adjectives. Its exploitation would allow to estimate the categories of lesser-known Points of Interest (POIs) or events. It can also allow to generate sentences and to answer questions that tourists may ask. The specificity of our approach lies in the Graph Convolutional Network (GCN)-based encoding of POIs with a graph of verbs, nouns, and adjectives and the estimation of POI categories using the proposed encoding. Experimental results show that our approach performs better than approaches in the literature.

DOI： 10.1109/ISKE60036.2023.10481229

Scopus

Web of Science

Web of Science

Publisher：CEUR Workshop Proceedings  

In this study, the target behavior of prosocial behavior was defined as garbage disposal (behavior of taking out the trash), and an intervention method using an inducement was proposed and evaluated. Specifically, we conducted an intervention in which the saturation of a trash can was visualized using a colored light that could be changed to red, yellow, and green. In addition, questionnaire surveys were administered before and after the intervention to 28 undergraduate and graduate students and faculty members, asking about their attitudes and feelings toward garbage disposal behavior, and personality traits. The results of a survey on the frequency of trash disposal be-fore and after the experiment showed that the time that the saturated trash can was left unattended was reduced by 81%. The questionnaire survey confirmed that the intervention increased the positive attitude toward trash disposal behavior, indicating that the intervention had an impact on trash disposal behavior. In addition, we were able to confirm the characteristics of individual personality traits and susceptibility to intervention.

Scopus

Publisher：2023 IEEE International Conference on Pervasive Computing and Communications Workshops and other Affiliated Events, PerCom Workshops 2023  

In this paper we present C-HAR, a low-cost,low-complexity, compressive measurement-based human activity recognitionframework for embedded devices, capable of simultaneously filtering andacquiring the sensor readings of a smartwatch's accelerometer and gyroscope atsub-Nyquist rates. A software simulation of the full system obtains an accuracyof 92.0% and 88.0% for the accelerometer-based and gyroscope-based systemsrespectively, obtained at a 5 Hz sample rate. A microcontroller implementationof the system's back-end obtains an accuracy of 90.3% and 87.1% for theaccelerometer-based and gyroscope-based systems respectively with a runtimeover twice as fast as that of a comparable baseline system.

DOI： 10.1109/PerComWorkshops56833.2023.10150315

Scopus

Web of Science

Language：Others   Publishing type：Research paper (scientific journal)   Publisher：IEEE Access  

Learning user attributes is essential for providing users with a service. In particular, for e-commerce portals which deal in variety of goods ranging from clothes to foods to home electronics, it is especially important to learn 'domain-independent' attributes such as age, gender, and personality that affect people's behavior across various domains of daily life (e.g., clothing, eating and housing) because these attributes can be used for personalization in diverse domains their service covers. Thus far, researchers have proposed approaches to learn user representation (UR) from user-item interactions, trying to embed rich information about user attributes in UR. However, very few can learn URs that are domain-independent without confounding them with domain-specific attributes (e.g., food preferences). This could consequently undermine the former's utility for personalizing services in other domains from which the URs are not learned. To address this, we propose an approach to learn URs that exclusively reflect domain-independent attributes. Our approach introduces a novel multi-layer RNN with two types of layers: Domain Specific Layers (DSLs) for modeling behavior in individual domains and a Domain Independent Layer (DIL) for modeling attributes that affect behavior across multiple domains. By exchanging hidden states between these layers, the RNNs implement the process of domain-independent attributes affecting domain-specific behavior and makes the DIL learn URs that capture domain-independence. Our evaluation results confirmed that the URs learned by our approach have greater utility in predicting behavior in the other domains from which these URs were not learned thereby demonstrating adaptability to various domains.

DOI： 10.1109/ACCESS.2022.3220781

Web of Science

Scopus

researchmap

Language：Others   Publishing type：Research paper (international conference proceedings)   Publisher：Proceedings - 2022 IEEE International Conference on Big Data, Big Data 2022  

コロナ禍における行動変容促進のために、金銭的インセンティブを動的に変化させた場合の影響、効果について、大規模、長期間に渡る社会実証を行った結果について報告した論文である。九州大学における混雑度回避行動を対象とした実験を７ヶ月以上行った結果、金銭的インセンティブを混雑度に応じて動的に変化させることで、行動変容に必要なインセンティブを約12%削減であることを示した。

DOI： 10.1109/BigData55660.2022.10020753

Scopus

researchmap

Other Link： https://dblp.uni-trier.de/db/conf/bigdataconf/bigdataconf2022.html#OchiaiSKIOYYOSA22

Language：Japanese  

カメラやマイクをミュートにしたオンライン会議で参加者の反応を知ることは困難である．対話相手からのフィードバックが少ない状況では，話者は話しづらさを感じ，聴者は会議への積極性が低下するといった課題がある．我々は，オンライン会議で不足している対話相手からのフィードバックを補填するために，参加者の表情認識から推定された感情を絵文字としてリアルタイムにフィードバックするシステム「REmotion」の開発を行っている．我々の目的は，本システムをオンライン会議で用いることで，話者の話しづらさを解消することと聴者の会議への積極性を向上させることである．そのために，参加者が対話相手を意識して表情を変化させることが重要だと考える．本論文では，開発したシステムを用いて実験を行い，参加者がカメラやマイクをミュートにして参加するオンライン会議で本システムを用いると，参加者が自分の表情を意識的に変化させることを明らかにした．

Language：English   Publishing type：Research paper (other academic)  

Function Estimation of Multiple IoT Devices by Communication Traffic Analysis

Language：Others   Publishing type：Research paper (scientific journal)   Publisher：IEEE Access  

It is necessary to optimize evacuation guidance to shelters in short evacuation time. The state-of-the-art method based on an idea of combinatorial optimization problems related to evacuees' locations and the capacities of nearby shelters has been developed, while it cannot mitigate the effect of congestion on roads/streets after evacuation starts. In this study, to cover this problem, we develop a new method that utilizes simulations for estimating the effect of congestion on roads/streets during evacuation and reassigning shelters to evacuees based on the simulation results. By iterating this step, our method derives the congestion-aware solutions for shelter selection that can realize more smooth evacuation. To evaluate our method, we conducted multi-agent simulations assuming a disaster situation in a sightseeing spot. Specifically, we examined a hypothetical case scenario involving the evacuation of 30,000 visitors from the Gion Festival. We compared the proposed method with conventional methods, such as the nearest shelter selection method and our previous method. We found that our proposed method reduced average and total evacuation time and congestion on roads compared to the conventional methods including the nearest shelter selection method and our previous method that only employs combinatorial optimization without estimating congestion. From this result, our idea of simulation-based congestion estimation has an impact of easing congestion during evacuation and preventing overcapacity of shelters at the same time. It shows the possibilities of help in developing congestion-aware evacuation strategies in emergency situations of crowded areas like huge cities or sightseeing spots.

DOI： 10.1109/ACCESS.2022.3194874

Web of Science

Scopus

researchmap

Language：Japanese  

路線バスは沿線の道路交通状況や天候など，外的要因により運行状況が左右されやすいため，遅延することが多い傾向にある．このため，他の交通機関への乗り換えが予定通りにできなくなる事例が観察されている．本研究では，路線バス走行履歴データから運行状況を解析し，乗り換え時間の設定を変えた場合にどのように乗り換えが成立するかを調査した．また，利用客の満足度を高めるような乗り換えを実現するために，改善が必要なダイヤを検出する指標の改良を検討した．

Language：English   Publishing type：Research paper (international conference proceedings)  

researchmap

Language：English   Publishing type：Research paper (other academic)   Publisher：UbiComp/ISWC 2022 Adjunct - Proceedings of the 2022 ACM International Joint Conference on Pervasive and Ubiquitous Computing and Proceedings of the 2022 ACM International Symposium on Wearable Computers  

Color-wall: Adaptive Color Filter to Reduce Digital Distractions during PC Work
As the Attention Economy has grown, numerous digital contents are made with the intention of capturing users' attention and spending more of their time by the use of attractive graphics and color psychology. In this paper, we propose a novel defense strategy, called a color-wall, to block digital distraction brought on by the consumption of visually seductive and addictive digital content. When it detects that a user is viewing media or using applications unrelated to the main task, it removes colors from the screen (gray-scaling) to make them less aesthetically pleasing and directs the user away from looking at them. We implemented a mac OS software as a proof-of-concept for the color-wall and conducted user testing (N=14) to assess the effectiveness of the proposed strategy. The evaluation results suggest that the color-wall is an effective strategy for reducing unwanted digital distractions such as browsing restricted content (YouTube, Twitter, etc.) while working on a PC.

DOI： 10.1145/3544793.3563403

Scopus

researchmap

Language：Others   Publishing type：Research paper (scientific journal)   Publisher：Sensors and Materials  

Many office workers are exposed to fluorescent light from mid-morning to after dark. The changes in the amount of light in the natural world, from bright during the day to dark at night, play an important role in maintaining a regular daily rhythm. In this study, we evaluated the effects of LED lighting equipped with a circadian rhythm control function, Lavigo, by comparing it with conventional fluorescent lighting on students' sleep quality and stress. Lavigo's Visual Timing Light (VTL) function adjusts the color and brightness of light according to the biological rhythm of about 24 h cycles. We measured the sleep state using Fitbit and collected subjective responses through a stress-checking questionnaire. A salivary amylase monitor was used to measure objective stress values from a medical perspective. No significant difference in stress values was observed during the experiment. However, only during the week that Lavigo was used, a certain subject showed a significant decrease in waking time during sleep, and another subject showed an increase in deep sleep. It was also found that these subjects tended to lead an irregular life, such as staying up until midnight.

DOI： 10.18494/SAM3952

Web of Science

Scopus

researchmap

Language：Others   Publishing type：Research paper (scientific journal)   Publisher：Sensors and Materials  

Maintaining a positive interaction is the key to a healthy and efficient meeting. Aiming to improve the quality of online meetings, we present an end-to-end neural-network-based system, named MeetingPipe, which is capable of quantitative microbehavior detection (smiling, nodding, and speaking) from recorded meeting videos. For smile detection, we build a neural network framework that consists of an 18-layer residual network for feature representation, and a selfattention layer to explore the correlation between each receptive field. To perform nodding detection, we obtain head rotation data as the key nodding feature. Then we use a gated recurrent unit followed by a squeeze-and-excitation mechanism to capture the temporal information of nodding patterns from head pitch angles. In addition, we utilize TalkNet, an active speaker detection model, which can effectively recognize active speakers from videos. Experiments demonstrate that with K-fold cross validation, the F1 scores of the smile, nodding, and speaking detection are 97.34, 81.26, and 94.90%, respectively. The processing can be accelerated with multiple GPUs due to the multithread design. The code is available at https://github.com/ humanophilic/MeetingPipe.

DOI： 10.18494/SAM3959

Web of Science

Scopus

researchmap

Language：Others   Publishing type：Research paper (other academic)   Publisher：Proceedings - IEEE Congress on Cybermatics: 2022 IEEE International Conferences on Internet of Things, iThings 2022, IEEE Green Computing and Communications, GreenCom 2022, IEEE Cyber, Physical and Social Computing, CPSCom 2022 and IEEE Smart Data, SmartData 2022  

The Internet is an effective tool for learners to gain new knowledge. Often, people use search engines (e.g., Google) rather than accessing websites directly. People have their search techniques to find specific information. In particular, people with domain knowledge tend to search more efficiently than novices. By understanding the gap between people with domain knowledge and novices, the novice can understand the path to becoming an expert. Therefore, in this study, we wanted to know what differences exist in search and programming behavior with and without domain knowledge. In this experiment, we asked a group with and without domain knowledge to solve ten programming problems and collected search logs (input knowledge) and compilation logs (output knowledge). Specifically, the first dataset consisted of 13 participants who had taken a university programming class. The second dataset consisted of 20 participants who had not taken a programming class and had no domain knowledge. We examined differences in search and compilation behavior based on participants' domain knowledge from this data. Since we observed a difference between each group when referring to the correlation coefficient, we performed a binary classification of novice and experienced participants using Random Forest, and achieved an average precision of 0.95, indicating that there were different trends in behavior with and without domain knowledge.

DOI： 10.1109/iThings-GreenCom-CPSCom-SmartData-Cybermatics55523.2022.00034

Scopus

researchmap

Other Link： https://dblp.uni-trier.de/db/conf/ithings/ithings2022.html#WatanabeMNAI22

Language：Japanese  

本研究では，行動変容を誘発する要素技術として，ゲーム等で利用される ''抽選'' がもたらす高揚感に着目し，ゲーム内で使用される抽選器を構成するどの要素が人の高揚感に影響を与えるかについて調査した．我々は，スロットを止めるボタンを押すという行為や手回しで抽選器を操作する行為が自己決定感を生み出し，さらに制御幻想を生じさせているのではないかと仮説を立て，抽選器として，手回し式の抽選器，自動でリールが停止するスロット，手動でリールを停止するスロットの 3 つの種類の抽選器を作成した．そして，実際のゲーム上で不特定多数の挙動を見るオンライン実験（アンケート回答者 428 名）と，表情筋や心拍など生理現象を定量的に計測するオフライン実験（被験者 1 名）を行った．その結果，スロットを自分で止めるという操作が制御幻想を生み出し，自己決定感が高揚感の向上に関係している可能性が示唆された．また，リーチによるニアミスなど当たりに近いハズレの存在も，ユーザの高揚感や表情の変化に影響を与える重要な要因であることが明らかとなった．

Language：Others   Publishing type：Research paper (other academic)   Publisher：Proceedings - 2022 12th International Congress on Advanced Applied Informatics, IIAI-AAI 2022  

Local events are important for the local people, tourists and local authorities. However, information about them is limited, or even absent on official websites. We propose an approach to automatically generate a data graph on local events via social network and mobile positioning data. This graph is constructed by the research results of syntactic and semantic data collection and will be used for an intelligent local event recommendation application. The coupling of these two research results makes it possible to discover little-known events or places that are not in the Linked Open Data and at the same time to connect them to it. An online experiment was set up to evaluate the performance of the techniques used in the study on semantic data which concerns the automatic ontology generation. The results allow us to conclude that the performance of the Named Entity Recognition and the choice of threshold score in automatic ontology generation depend on the categories of detected words types.

DOI： 10.1109/IIAIAAI55812.2022.00034

Scopus

researchmap

Other Link： https://dblp.uni-trier.de/db/conf/iiaiaai/iiaiaai2022.html#RajaonarivoMA22

Language：English  

A study on estimating the accurate head IMU motion from Video

Language：English  

A Study on 3D Gaze Recognition from the Environment Side by Multiple Optical Sensors

Language：Japanese  

近年，一般家庭にも IoT デバイスが普及し，リモコンや照明，ドアロック，コンセントなど様々な機能を持つ IoT デバイスが販売され，様々な場面で活用されている．現在の IoT デバイスは動作がブラックボックスであり，IoT デバイスがユーザの意図しない通信を行っていた場合に気づく術がない．そこで，我々は IoT デバイスがどのような通信を行っているかを検知し，それをもとに適切な通信のみ許可することができるアクセス制御の機能と IoT デバイスがどのような通信を行っているか可視化することでユーザが IoT デバイスの動作状況を理解することを可能にする機能を持つシステム (IoT 活動量計) の実現を目指している．その実現のため，本研究では，IoT デバイスの通信トラヒックを分析し，どの機能が使われているかを推定する手法を提案し，4 種別の IoT デバイス各 2 機種ずつの計 8 機種の 8 種類の機能の通信トラヒックを収集し，分類を行った．その結果，デバイスの機種と実行した機能の組み合わせの 16 通りで分類した場合，88% の精度で機能を推定できることを確認した．また，実行した機能のみの 8 通りで分類した場合，69% の精度で機能を推定できることを確認した．

Language：Japanese  

オンライン購買システムは，ページ遷移やクリックの履歴から購買者が商品の購買に至るまでの過程を記録することができる．これにより，興味がある商品や比較検討した商品がわかり，購買者のニーズに近く購買意欲を高めるような商品推薦が可能である．一方，実店舗においては商品の購買以外の情報は，取得と活用の両面における難しさからあまり記録されていない．そのため，購買者がどのような商品に興味を持っており何で悩んでいるのかがわからないという問題がある．こうした問題を解決するために，我々はLiDAR（Light Detection And Ranging）からの点群データとDepth（深度）カメラからの視線情報を用いた購買時における候補商品推定システムを構築し，その初期評価を行った．推定の上では，興味のある商品を購買候補商品とした．評価の結果，購買時の視線は候補商品と強く結びつき，視線の情報は候補商品を推定する上で非常に重要な特徴となり得ることがわかった．一方で，視線の評価を踏まえ点群データを使用した候補商品の予測を行った結果，予測対象の人物を含んだ状態で学習したモデルは49%と12クラス分類であることを考えると一定の精度を出すことができた．一方でLOPOでは精度が最大22%と，モデルの一般化はまだまだ課題が残ることがわかった．今後は，データ数を増やし，モデルをより適したものに変更・改善していく必要がある．

Language：Japanese  

新型コロナウイルスの流行による外出制限を受けて料理動画の需要が高まっているが，料理中に動画を視聴する際，動画再生端末を操作する必要がある．そこで本研究ではユーザの料理工程に合わせて料理動画をループ再生する手法を提案する．ユーザが行っている料理工程を動画内でループ再生することで動画再生端末を操作することなく料理動画を視聴することが可能となる．本稿では，この実現に向けた第一歩として料理音から料理工程を認識する手法について述べる．ユーザの料理音を取得し，既存の料理音データセットで作成した機械学習モデルに入力することで料理工程を認識する．実環境ではテレビの音や BGM が混ざる可能性があるためバックグラウンドノイズの影響について検討した結果も報告する．一般家庭で録音した料理音データを kitchen20 データセットで作成した分類モデルで初期的評価を行った結果，3 種類の料理工程を正解率 0.84 で分類できることを確認した．また料理工程が頻繁に変わらないことを考慮して平滑化処理を行った結果も示す．

Language：Others   Publishing type：Research paper (scientific journal)   Publisher：Sensors and Materials  

In this paper, we propose a framework called an IoT activity tracker to ensure the safe and secure use of IoT devices around the home. The IoT activity tracker identifies the types of IoT devices and their triggering functions by traffic pattern analysis, and provides information about which IoT devices are running in the home. Also, it allows users to temporarily or permanently block the network of a particular IoT device from their smartphone if the user notices unauthorized activity of the IoT device. Furthermore, the system can be linked to sensor-based behavior recognition. For example, an IoT-based smart lock can automatically switch the communication permissions between those when the user is out and those when the user is at home depending on its status. In this paper, we report the results of a proof of concept experiment on our proposed system, which consists of a router with the proposed function and a cloud server that interfaces with users.

DOI： 10.18494/SAM3901

Scopus

researchmap

Language：Japanese   Publisher：The Japanese Society for Artificial Intelligence  

Difficulties in continuous physiological sensing in-the-wild
To realize ubiquitous services based on physiological sensing, it is important to assume that commercially available sensors will be used in daily life. Collecting physiological data from ordinary people for a long period of time in-the-wild is completely different from performing highly accurate physiological sensing using dedicated sensors in a controlled laboratory environment. In this presentation, we will share the various problems we have experienced and clarify the points that need to be considered in order to achieve stable sensing in a real environment.

DOI： 10.11517/pjsai.jsai2022.0_3i3os5a01

CiNii Research

researchmap

Language：Japanese   Publisher：The Japanese Society for Artificial Intelligence  

Quantifying a Multi-person Meeting based on Multi-modal Micro-behavior Analysis
In this paper, we present an end-to-end online meeting quantifying system, which can exactly detect and quantify three micro-behavior indicators, speaking, nodding, and smile, for online meeting evaluation. For active speaker detection (ASD), we build a multi-modal neural network framework which consists of audio and video temporal encoders, audio-visual cross-attention mechanism for inter-modality interaction, and a self-attention mechanism to capture long-term speaking evidence. For nodding detection, based on the WHENet framework proposed in the research field of head pose estimation (HPE), we can estimate the head pitch angles as the nodding feature. Then we build a gated recurrent unit (GRU) network with squeeze-and-excitation (SE) module to recognize nodding movement from videos. Finally, we utilize a Haar cascade classifier for smile detection. The experimental results using K-fold Cross Validation show that the F1-score of each detection module achieves 94.9%, 79.67% and 71.19% respectively.

DOI： 10.11517/pjsai.jsai2022.0_1p1gs1004

CiNii Research

researchmap

Language：Japanese   Publisher：The Japanese Society for Artificial Intelligence  

Construction and evaluation of a human-in-the-loop video annotation system for nodding during meetings
Annotation is one of the time consuming manipulation for human to do. In our previous work, we made a nodding recognition model using annotated video data made by an annotator. We find that consumption of annotation time and the work load require for the annotator is huge. In this work we come up with the idea of human-in-the-loop video annotation system, which the machine leaning model create annotation result before the annotator finalize the annotation result.With this system we evaluated how average time consumed in annotation is reduced.

DOI： 10.11517/pjsai.jsai2022.0_3h3os12a03

CiNii Research

researchmap

Language：English   Publisher：(NPO)日本脳神経血管内治療学会  

目的:脳血管内治療を行う際に,どこを注視しながら操作を行うかは重要である.アイトラッカーを用いた視線解析により治療経験に応じた脳血管内治療中の視線の違いを検出可能か検証した.方法:被験者は当院で脳血管内治療に従事する治療経験歴の異なる7人の医師とした.治療経験によりExpert群,Trainee群,Novice群に分け,各群での脳血管内治療中の視線の違いを検討した.治療操作には人体血管モデルを利用し,脳動脈瘤に対しbiplane透視下でマイクロカテーテルの動脈瘤内への誘導,コイル挿入を行い,操作中の視線をアイトラッカーで抽出した.結果:マイクロカテーテルの動脈瘤内への誘導では,Expert群,Trainee群が主に正面像を注視して操作しているのに対し,Novice群では側面像とモニター外を見ている時間が長かった.コイル挿入ではExpert群が正面像をメインの画像として注視しサブ画像として側面像としてバランス良く利用していた.一方でTrainee群はほぼ正面像のみを注視し,Novice群はモニター外を見る時間が他の2群に比べ長かった.結論:治療経験による脳血管内治療中の視線の違いがアイトラッカーを用いて解析することが可能であった.今後の脳血管内治療の教育に活用できると期待される.(著者抄録)

Language：Japanese  

Language：Japanese  

Language：Others   Publishing type：Research paper (other academic)   Publisher：IEEE  

In this paper, we propose a method for estimating psychological states such as hesitation and interest during survey responses based on touch operation behavior (touch log). In addition, we propose two enhanced web survey RUIs (Response User Interfaces), (a) a slidebar type RUI and (b) a magnifier type RUI, to expand the operational differences due to hesitation and interest for more accurate estimation. To evaluate the proposed RUIs, we implemented Augmented Web Survey (AWS) as an extension of LimeSurvey and conducted two experiments. Responses were collected from 8 subjects in the experiment to evaluate (a) a slidebar type RUI and from 21 subjects to evaluate (b) a magnifier type RUI. We measured the degree of hesitation by evaluating the subjective hesitation to the questions on a 5-point scale. For the measurement of the degree of interest, the participants were asked to select their top two favorite images among the six images, and the true value was taken as the top two. As a result of comparing the operation logs of the conventional RUI and the slidebar type RUI, it was confirmed that there was a significant difference depending on the degree of hesitation. In addition, more magnification operations were obtained for the magnifier type RUI than the standard pinch-out operation suggesting that it may help estimate psychological states such as hesitation.

DOI： 10.1109/LifeTech53646.2022.9754964

Scopus

researchmap

Other Link： https://dblp.uni-trier.de/db/conf/lifetech/lifetech2022.html#NakagawaAN22

Language：Japanese   Publishing type：Research paper (scientific journal)  

Language：Others   Publishing type：Research paper (other academic)  

In this study, we propose a behavior change support system that encourages people to use stairs instead of elevators. Using the stairs is not only good for one's health, but nowadays it also plays a role in dense avoidance from the perspective of COVID-19. Although many people are waiting for the elevator, we thought that if we knew how many people were on the elevator, more people could change their behavior. However, there are few buildings where the number of people in the elevator is displayed on each floor. So, we have developed a headcount measurement system in the cargo and a visualization system, those can be easily appended to the current elevator. As a method to measure the number of people in an elevator, we propose a method to detect BLE (Bluetooth Low Energy) signals transmitted from the terminals of elevator users who have installed COCOA, an application for confirming contact with the COVID-19 in Japan, and to measure in real time the number of detected BLE signals with a received signal strength (RSSI) exceeding a certain value. We propose a method to measure the number of BLE signals detected in real time. In this paper, we have designed and developed the system, verified the accuracy of the detection of the continuous operation time and the number of passengers, and estimated the behavior pattern and waiting time of elevator users using the system. The evaluation of the transformation of the decision making of the elevator users brought about by this system is out of the scope of this paper and will be discussed in the future.

Language：Japanese  

オンラインアンケートにおいて，なるべく楽に早くアンケートタスクを完了しようとする「Satisficing（努力の最小限化）」という態度が調査結果の信頼性を低下させる問題がある．この問題に対し，回答時間が明らかに短い回答を除外するなどの措置が考案されてきたが，回答時間だけでは完全に除外することができない．そこで我々は，回答時間だけでなく，回答中の画面操作を観測することで，より高精度にSatisficingが検出可能になるのではないかという仮説を立てた．世界中で利用されているオープンソースアンケートシステムLimeSurveyの拡張機能として，回答中の画面操作ログ（回答操作ログ）を記録できるプラグインを開発し，本システムを用いてクラウドソーシング上でアンケート実験を行った．実験は，回答者は各自のスマートフォンを用いて，リ
ッカート形式と自由記述形式の質問から成るアンケートに回答してもらうものであり，アンケートの途中にはSatisficingを検出するために考案されている既存の質問を挿入した．1000人に対して回答を依頼し，収集した回答内容および回答操作ログから，
Satisficingを表現する可能性が考えられる特徴量を生成し，特徴量ごとにSatisficingである群とSatisficingでない群の母平均の差の検定を行った．結果として，既存の特徴量であるテキストの文字数，連続同一回答数，中間回答数に加え，開発したプラグインでのみ取得可能なスクロール速度，選択肢の変更回数，リッカート形式の回答時間などに統計的有意差があることを明らかにした．また，スクロール速度やテキストの変更回数に関しては，絶対値では差がない一方で，回答者ごとのベースラインとの差では有意な差が確認できた．
identifier:2189-0374

Language：Japanese  

type:Journal Article
都市部への人口集中と自動車の利用の増加は，交通渋滞を引き起こす原因の一つとなっている．ユーザが交通渋滞を避けるためには，車車間通信(V2V) によるリアルタイムで交通情報が共有できるシステムが有用であると考えられる．しかし，V2V通信の問題点として，建造物や車両の影響により，見通しが取れない環境下での長距離伝送は困難であることが挙げられる．本研究では，低コスト・低消費電力で広域の通信を実現するため，Private LoRaを用いた車車間情報共有のためのデータ転送中継方式を提案する．提案するデータ中継転送方式の有効性を実証実験により検証する． / Population concentration in urban areas and the increase in the use of automobiles are causes of traffic congestion jams. The system sharing traffic information in real-time by vehicle-to-vehicle (V2V) communication is useful for users to avoid traffic jams. However, one of the problems with V2V communication is that it is difficult to transmit over long distances in an environment where visibility is limited due to the influence of buildings and moving vehicles. This paper proposes a data relay scheme for inter-vehicle information sharing using private LoRa with long-range communication. We verify the effectiveness of our proposed data relay scheme through demonstration experiments.
電子情報通信学会 ネットワークシステム研究会（NS）, 2022年3月10日-11日, オンライン開催
source:https://www.ieice.org/ken/paper/20220311TCjl/

Language：Others   Publishing type：Research paper (other academic)   Publisher：2022 IEEE International Conference on Pervasive Computing and Communications, PerCom 2022  

複数のエナジーハーベスト素子を組み合わせ，発電量の変化をセンサとして活用することで，バッテリーレスなライフロギングを実現した．

DOI： 10.1109/PerCom53586.2022.9762376

Web of Science

Scopus

researchmap

Other Link： https://dblp.uni-trier.de/db/conf/percom/percom2022.html#AritaNIA22

Language：Others   Publishing type：Research paper (other academic)   Publisher：Springer International Publishing  

To promote the face-to-face communication reduced by COVID-19, we proposed and implemented a context-aware Slack chatbot based on Cyber-Physical sensing that helps colleagues meet more often in the same place. Our system periodically collects the user’s internal context through Slack (cyber sensing) and uses small BLE beacons distributed to colleagues and beacon scanners installed in a laboratory to sense physical attendance (physical sensing). In addition, the system notifies the user of recommended actions, such as lunch or coffee break, depending on the context determined by the Cyber-Physical sensors. We deployed the proposed system in a laboratory environment and conducted an initial experiment for six weeks. Experimental results confirmed that our system can encourage serendipitous face-to-face communication during periods when the frequency of attending school and going to work dropped due to COVID-19. It was also found that in an environment such as a laboratory, where a certain level of trust has already been established, the openness of the collected information can further motivate users to participate in the system.

DOI： 10.1007/978-3-030-98438-0_18

Web of Science

Scopus

researchmap

Other Link： https://dblp.uni-trier.de/db/conf/persuasive/persuasive2022.html#TanakaMNA22

Publishing type：Research paper (international conference proceedings)  

In this study, we propose a behavior change support system that encourages people to use stairs instead of elevators. Using the stairs is not only good for one's health, but nowadays it also plays a role in dense avoidance from the perspective of COVID-19. Although many people are waiting for the elevator, we thought that if we knew how many people were on the elevator, more people could change their behavior. However, there are few buildings where the number of people in the elevator is displayed on each floor. So, we have developed a headcount measurement system in the cargo and a visualization system, those can be easily appended to the current elevator. As a method to measure the number of people in an elevator, we propose a method to detect BLE (Bluetooth Low Energy) signals transmitted from the terminals of elevator users who have installed COCOA, an application for confirming contact with the COVID-19 in Japan, and to measure in real time the number of detected BLE signals with a received signal strength (RSSI) exceeding a certain value. We propose a method to measure the number of BLE signals detected in real time. In this paper, we have designed and developed the system, verified the accuracy of the detection of the continuous operation time and the number of passengers, and estimated the behavior pattern and waiting time of elevator users using the system. The evaluation of the transformation of the decision making of the elevator users brought about by this system is out of the scope of this paper and will be discussed in the future.

Scopus

researchmap

Language：Japanese   Publishing type：Research paper (scientific journal)  

researchmap

Language：Japanese   Publisher：[出版社不明]  

Estimating Phone Holding Style through Analysis of Touch Logs
ユーザがどのようにスマートフォンを持ち，どの指で操作しているかという情報（操作形態）が分かれば，様々な新しい応用，たとえば，操作形態に適したUI（ユーザインターフェス）への動的切替えや，ながらコンテキスト（スマートフォン操作時に行われている別行動，歩行，喫煙，食事など）の推定など，への道が拓ける可能性がある．本論文では，タッチ操作によって出力されるログに着目し，タッチやスワイプなどの操作領域や頻度からどのような操作形態かを推定する方法を提案する．そのために，Android上でアプリ横断的にタッチ操作を取得可能なTouchAnalyzerから得られた情報を特徴量とし，機械学習手法により代表的な8通りの操作形態を推定する手法を構築する．16名の実験参加者による評価実験を通じて，提案手法によって操作形態をおおむね90&#37;のF値で推定できることが分かった．
In this paper, we propose a method for estimating the phone holding style that shows hand grasp and finger placement while using a smartphone (e.g., either by both hands or one hand or by putting it on the desk) through analysis of screen touch behavior logs. Knowing the phone holding style in real-time will open the door to various new applications such as dynamically providing a suitable interface (e.g., layouts of buttons) and estimating “side-contexts” like walking, smoking and drinking while using a smartphone. In this paper, focusing on the logs that Android operating systems generate when touch behavior happens, we propose a method to estimate the way of holding a smartphone by area (in the screen) and frequency of touch behavior. We have developed a tool called TouchAnalyzer that identifies touch behavior types like swipe, rotate, pinch, multi-touch, etc. from the device-dependent logs of Android OS. By using the touch behavior type logs that TouchAnalyzer outputs, we have developed an algorithm that can recognizre eight representative phone holding styles. We conducted evaluation experiments with 16 participants and confirmed that the proposed method achives about 90&#37; of F-score in estimating the phone holding style.

DOI： 10.20729/00215710

CiNii Books

CiNii Research

J-GLOBAL

researchmap

CiNii Research

CiNii Research

Language：Japanese   Publisher：The Japanese Psychological Association  

<p>本シンポジウムは，第66回大会から継続して企画されているシンポジウムおよびワークショップ「職場のストレスとメンタルヘルス」の第20回目に相当する。</p><p>わが国では，雇用・労働環境の変化，価値観や働き方の多様化，デジタル化が進展し，社会や働き方が大きく変化している。一方，一人ひとりの労働者も，コロナ禍で働き方の変化に直面する中，働く意義を再考し，より充実した働き方を実現するための方策を模索し始めている。つまり，現在は，社会全体が新しい労働観や働き方へと向かう転換期にあると言える。</p><p>そこで，本シンポジウムでは，これからの職場のメンタルヘルスについて，特に労働観の転換と朗働の実現に注目して検討する。具体的には，（1）朗働/幸福の心理生物的メカニズム，（2）組織内の行動，（3）勤怠データを用いた働き方と休み方，（4）情報技術を用いた行動の測定，の4点を取り上げ，「職業生活においてどのようにウェルビーイングを促進するか？」という課題について検討したい。</p>

DOI： 10.4992/pacjpa.86.0_ss-038

CiNii Research

Publisher：Sensors and Materials  

The rapid proliferation of mobile devices with various sensors has enabled participatory mobile sensing (PMS). Several PMS platforms suffer from open issues including the limited use of their functions to a specific scenario/case and the necessity of technical knowledge for organizers. This paper proposes a novel PMS platform named ParmoSense for easy and flexible data collection. To reduce the burden on both organizers and participants, we employ two novel features. First, essential PMS functions implemented as modules can be easily chosen and combined for sensing in different scenarios. Second, the scenario-based description feature allows organizers to easily and quickly prepare a new instance of PMS and enable people to easily participate in the sensing. It also provides multiple functions to motivate participants for sustainable operation. Through a performance comparison with existing PMS platforms, we confirmed that ParmoSense shows the best cost performance in terms of the workload for preparation and the variety of functions. In addition, to evaluate the availability and usability of ParmoSense, we conducted 19 case studies over four years with ordinary citizens. As the result of a questionnaire survey carried out during the case studies, we confirmed that ParmoSense can be easily operated by ordinary citizens without technical skills.

DOI： 10.18494/SAM3961

Web of Science

Scopus

Web of Science

Publisher：CEUR Workshop Proceedings  

In this study, we propose a behavior change support system that encourages people to use stairs instead of elevators. Using the stairs is not only good for one's health, but nowadays it also plays a role in dense avoidance from the perspective of COVID-19. Although many people are waiting for the elevator, we thought that if we knew how many people were on the elevator, more people could change their behavior. However, there are few buildings where the number of people in the elevator is displayed on each floor. So, we have developed a headcount measurement system in the cargo and a visualization system, those can be easily appended to the current elevator. As a method to measure the number of people in an elevator, we propose a method to detect BLE (Bluetooth Low Energy) signals transmitted from the terminals of elevator users who have installed COCOA, an application for confirming contact with the COVID-19 in Japan, and to measure in real time the number of detected BLE signals with a received signal strength (RSSI) exceeding a certain value. We propose a method to measure the number of BLE signals detected in real time. In this paper, we have designed and developed the system, verified the accuracy of the detection of the continuous operation time and the number of passengers, and estimated the behavior pattern and waiting time of elevator users using the system. The evaluation of the transformation of the decision making of the elevator users brought about by this system is out of the scope of this paper and will be discussed in the future.

Scopus

Language：Others   Publishing type：Research paper (other academic)  

The study of mobile positioning data makes it possible to detect whether an event has happened at a particular place during a given period. However, determining the nature and details of the event is a challenge, especially if the event is not widely known, as is the case for local events. We propose an approach to determining the nature of local events by generating an ontology in a completely automatic way from social network data and data on people's movements and by querying this generated ontology. This approach uses entity discovery techniques, filtering systems and information enrichment via Open Data, as well as a system for matching discovered entities and ontology elements. Evaluation via a survey allowed us to validate approximately that the information presented in the ontology is reliable, makes sense and answers our questions.

DOI： 10.1145/3486622.3493933

Language：Others   Publishing type：Research paper (scientific journal)  

As society grows ever more interconnected, the need for sophisticated signal processing and data analysis techniques becomes increasingly apparent. This is particularly true in the field of intelligent transportation systems (ITSs), where various sensing applications generate data at an exponential rate. In this paper, we present C-AVDI, a compressive measurement-based acoustic vehicle detection and identification architecture capable of extracting information from vehicle audio signals while sampling at sub-Nyquist rates. In addition, we further reduce the overall complexity by performing any necessary signal filtering during the acquisition process, removing the need for a separate filtering stage in the system's front-end. Our results obtained from data collected under a range of weather conditions present an accuracy of 80&#37; with a back-end analog-to-digital converter (ADC) sample rate of 3kHz, with initial results from a microcontroller (MCU) implementation of our proposed system presenting an accuracy of 72&#37;.

DOI： 10.1109/ACCESS.2021.3132061

Language：Others  

Development of Summarization System for Electronic Medical Record using Artificial Intelligence

Language：Others   Publishing type：Research paper (other academic)  

Occupational health issues such as depressions have become a serious problem in recent years. There is an urgent necessity for the early preventive detection of depression rather than a yearly survey. Toward detecting depression signs, this paper statistically analyzes the relationship between occupational health indicators collected by questionnaires and sleep status (sleeping condition and heart rate data) during bedtime measured by a common wearable device.

DOI： 10.23919/ICMU50196.2021.9638882

Language：Others   Publishing type：Research paper (other academic)  

This study aims to estimate the work engagement (WE) of office workers using biological data related to their daily activities obtained from wearable devices, with the goal of providing appropriate mental and physical health support to each individual. We collected daily heart rate data from wearable devices worn by 60 office workers in five Japanese companies for 2-3 weeks. Their daily WE was measured using the state-of the-art utrecht work engagement scale (UWES) questionnaire. We performed two types of analysis on the collected data using machine learning methods. First, the classification of binary WE levels (high or low), which showed a leave-one-person-out (LOPO) cross-validation F1 value of 0.522. Second, we classified whether WE decreased compared to the previous day, which showed a LOPO cross-validation F1 value of 0.663.

DOI： 10.23919/ICMU50196.2021.9638884

Language：Others   Publishing type：Research paper (other academic)   Publisher：Lecture Notes of the Institute for Computer Sciences, Social-Informatics and Telecommunications Engineering, LNICST  

In this study, we propose a room-layout-based appliance control for voice user interfaces (VUIs), such as smart speakers. VUI-based appliance control requires a control command including which device to do what. However, we often experience an ambiguous target problem: the control target device in a control command is ambiguous because an ambiguous room name and demonstrative words are frequently used to specify the target device. To address this problem, we utilized a room layout to estimate the control target. A user implicitly aims to control devices in a room where they are. Therefore, we estimate the room where the user is now based on the room layout, which is estimated on a smart speaker, to determine the control target. In this study, we present the design of a room-layout estimator as the first step toward room-layout-based appliance control. The experimental evaluations conducted in our 1-bedroom smart house revealed that our room-layout estimator estimates room directions and room types with accuracies of 0.850 and 0.714, respectively.

DOI： 10.1007/978-3-030-94822-1_2

Web of Science

Scopus

Language：Others   Publishing type：Research paper (other academic)  

The recent COVID-19 pandemic has made it extremely important to avoid crowded environments. In light of this, we are developing a congestion measurement system utilizing IoT technology, in which the congestion state of a given location is estimated by counting the number of WiFi Probe Request messages and BLE exposure notification messages sent from smartphones and laptops. Our congestion measurement system, however, suffers from an unstable measurement problem because the number of WiFi and BLE messages is highly dependent on the environment in which our system is deployed. This paper presents a congestion measurement system employing an automatic parameter adjuster that accurately estimates the number of people in various target locations. The parameters are automatically adjusted based on the seating capacity and area size of the target environment, as well as information gathered from a sensor which is designed to collect and analyze WiFi and BLE messages collected in a university cafeteria.

DOI： 10.23919/ICMU50196.2021.9638886

Language：Others   Publishing type：Research paper (other academic)   Publisher：Lecture Notes of the Institute for Computer Sciences, Social-Informatics and Telecommunications Engineering, LNICST  

There is an abundant and constantly growing amount of information that can be retrieved from online resources. Moreover, the access to such resources is becoming more and more convenient. Yet, finding the exact needed information is not easy, especially for programming search queries. In this paper, we present TrackThinkTS, a privacy-aware browser extension. It tracks users’ behaviors when navigating the web. The extension logs various user actions related to tab management, search query, browsing, and clipboard management. The extension is built with a privacy-first mindset. In fact, the users have full control over the registered logs, they can manage, update and export the logs in a completely transparent way. The vision behind this work is twofold. On one hand, we aim to investigate the web search behavior of programming students and detect patterns of a successful search. On the other hand, the objective is to build a knowledge base that will serve as a course supplement for programming students. Therefore, the proposed extension in this paper is one of the building blocks of the whole system. Data collected from this extension will be also synchronized with log data coming from an online IDE used by programming students during the experiment phase.

DOI： 10.1007/978-3-030-94822-1_51

Web of Science

Scopus

Language：Others   Publishing type：Research paper (scientific journal)   Publisher：The Japanese Society for Neuroendovascular Therapy  

<p><b>Objective</b>: In various fields, differences in eye-gazing patterns during tasks between experts and novices have been evaluated. The aim of this study was to investigate gazing patterns during neuro-endovascular treatment using an eye-tracking device and assess whether gazing patterns depend on the physician’s experience or skill.</p><p><b>Methods</b>: Seven physicians performed coil embolization for a cerebral aneurysm in a silicone vessel model under biplane X-ray fluoroscopy, and their gazing patterns were recorded using an eye-tracking device. The subjects were divided into three groups according to experience, highly experienced (Expert) group, intermediately experienced (Trainee) group, and less experienced (Novice) group. The duration of fixation on the anterior–posterior (AP) view screen, lateral (LR) view, and out-of-screen were compared between each group.</p><p><b>Results</b>: During microcatheter navigation, the Expert and Trainee groups spent a long time on fixation to AP, while the Novice group split their attention between each location. In coil insertion, the Expert group gazed at both the AP and the LR views with more saccades between screens. In contrast, the Trainee group spent most of their time only on the AP view screen and the Novice group spent longer out-of-screen.</p><p><b>Conclusion</b>: An eye-tracking device can detect different gazing patterns among physicians with several experiences and skill levels of neuroendovascular treatment. Learning the gazing patterns of experts using eye tracking may be a good educational tool for novices and trainees.</p>

DOI： 10.5797/jnet.oa.2021-0053

Web of Science

PubMed

CiNii Research

Language：Others   Publishing type：Research paper (other academic)   Publisher：Smart Innovation, Systems and Technologies  

In recent years, many organizations have prioritized efforts to detect and treat mental health issues. In particular, office workers are affected by many stressors, and physical and mental exhaustion, which is also a social problem. To improve the psychological situation in the workplace, we need to clarify the cause. In this paper, we conducted a 14-day experiment to collect wristband sensor data as well as behavioral and psychological questionnaire data from about 100 office workers. We developed machine learning models to predict psychological indexes using the data. In addition, we analyzed the correlation between behavior (work content and work environment) and psychological state of office workers to reveal the relationship between their work content, work environment, and behavior. As a result, we showed that multiple psychological indicators of office workers can be predicted with more than 80% accuracy using wearable sensors, behavioral data, and weather data. Furthermore, we found that in the working environment, the time spent in “web conferencing”, “working at home (living room)”, and “break time (work time)’ had a significant effect on the psychological state of office workers.

DOI： 10.1007/978-981-19-0361-8_1

Web of Science

Scopus

Language：Others   Publishing type：Research paper (scientific journal)  

The emergence of various types of commercial cameras (compact, high resolution, high angle of view, high speed, and high dynamic range, etc.) has contributed significantly to the understanding of human activities. By taking advantage of the characteristic of a high angle of view, this paper demonstrates a system that recognizes micro-behaviors and a small group discussion with a single 360 degree camera towards quantified meeting analysis. We propose a method that recognizes speaking and nodding, which have often been overlooked in existing research, from a video stream of face images and a random forest classifier. The proposed approach was evaluated on our three datasets. In order to create the first and the second datasets, we asked participants to meet physically: 16 sets of five minutes data from 21 unique participants and seven sets of 10 min meeting data from 12 unique participants. The experimental results showed that our approach could detect speaking and nodding with a macro average f1-score of 67.9&#37; in a 10-fold random split cross-validation and a macro average f1-score of 62.5&#37; in a leave-one-participant-out cross-validation. By considering the increased demand for an online meeting due to the COVID-19 pandemic, we also record faces on a screen that are captured by web cameras as the third dataset and discussed the potential and challenges of applying our ideas to virtual video conferences.

DOI： 10.3390/s21175719

Language：Japanese  

Population Flow Distribution Statistics which Capture the Flow of Persons and Change the World：Promotion of EBPM in Tourism - The Development of a System Based on Large-scale Ad Location Data -
緊迫化する地方自治体の財政への対応として，限られた財源を有効に活用するために，施策成果をより正確に分析し，見直していくEBPM（客観的証拠に基づく政策立案）の推進が求められてきた．しかしながら，現在多くの地域では施策の効果検証や分析まで人手が回らず，また外部委託が必要でお金と時間が足りないなどの課題を抱えており，本格導入には至っていない．そうした状況を打破すべく，国内最大規模の広告位置情報ビッグデータを活用して観光拠点単位で人流の検出と集計を担う政策立案評価システムを開発，自治体職員がコロナ禍の人流抑制などを即時かつ安価に把握で環境を整えることで観光分野におけるEBPMの推進を目指す取り組みを紹介する．

DOI： 10.20729/00212203

Language：English   Publishing type：Research paper (other academic)  

IoT devices are currently used in various situations, with the number of applications steadily increasing over the past few years. These IoT devices are connected to outside networks, and as such are often targets of hacking attempts and associated with private user information leaks. The end user is often unaware of how these devices operate, and as a result is unable to notice any unauthorized communication. In response, we aim to create a system to visualize the activity of an IoT device (IoT activity monitor). To create this system, we analyze the network traffic of an IoT device and propose a way to estimate which function it uses. Using Wire-Shark, a packet capture software, we evaluate the data transmitted from the Amazon Echo Spot, Amazon Echo Dot and Amazon Echo Flex, which are smart speakers. We confirmed that we can estimate 8~10 kinds of called functions with 76.1&#37;, 89.8&#37; and 85.2&#37; accuracy for each smart speaker respectively.

DOI： 10.1145/3412841.3441951

Language：Others   Publishing type：Research paper (other academic)  

When working from home, self-management becomes of paramount importance due to the absence of a boss or colleagues. As a result, individuals tend to waste time surfing the internet and playing with our smartphones. We propose a wrist-worn sensor-based system that identifies whether a desk worker is working or not for self-management and productivity. Our main hypothesis is that the identification of the various tasks that occur during desk work, such as using computers, reading books, manipulating a smartphone, and writing, can be simply distinguished by the direction of the palm. In this paper, to verify our hypothesis, we measure various tasks with the wrist-worn sensor attached to clarify the relationship between hand orientation and each task. At the same time, we develop a machine learning-based classifier to distinguish between the states of’working’ and’not-working’ using the obtained hand orientation data. We performed 10-fold cross-validation and Leave-One-Person-Out validation and we found that it was possible to distinguish whether a desk worker is working or not with an F1-value of 0.8 or higher.

DOI： 10.1109/LifeTech52111.2021.9391950

Language：Others   Publishing type：Research paper (scientific journal)  

Rapid proliferation of mobile devices with various sensors have enabled
Participatory Mobile Sensing (PMS). Several PMS platforms provide multiple
functions for various sensing purposes, but they are suffering from the open
issues: limited use of their functions for a specific scenario/case and
requiring technical knowledge for organizers. In this paper, we propose a novel
PMS platform named ParmoSense for easily and flexibly collecting urban
environmental information. To reduce the burden on both organizers and
participants, in ParmoSense, we employ two novel features: modularization of
functions and scenario-based PMS system description. For modularization, we
provide the essential PMS functions as modules which can be easily chosen and
combined for sensing in different scenarios. The scenario-based description
feature allows organizers to easily and quickly set up a new participatory
sensing instance and participants to easily install the corresponding scenario
and participate in the sensing. Moreover, ParmoSense provides GUI tools as well
for creating and distributing PMS system easily, editing and visualizing
collected data quickly. It also provides multiple functions for encouraging
participants' motivation for sustainable operation of the system. Through
performance comparison with existing PMS platforms, we confirmed ParmoSense
shows the best cost-performance in the perspective of the workload for
preparing PMS system and varieties of functions. In addition, to evaluate the
availability and usability of ParmoSense, we conducted 19 case studies, which
have different locations, scales, and purposes, over 4 years with cooperation
from ordinary citizens. Through the case studies and the questionnaire survey
for participants and organizers, we confirmed that ParmoSense can be easily
operated and participated by ordinary citizens including non-technical persons.

Language：Others   Publishing type：Research paper (other academic)  

Wi-Fi signal based detection is widely implemented in indoor action detection because of its low-cost and easy implementation. But it is still rarely used in equipment vibration detection. Moreover, it is hard to detect multiple targets where we need to monitor multiple equipments' vibration state such as in the factory environment. In this paper, we propose a wireless based vibration sensing method using Wi-Fi for factory equipment fault detection. First, we use CSI amplitude data to distinguish sensing target equipments. Then, we apply an anomaly detection method to detect faulty machine operation. We conducted initial experiments to validate the feasibility of our proposed fault detection method. The experimental results show that our method detected abnormal situations with an accuracy of 100&#37;, while 10&#37; of normal situations were mistakenly recognized as abnormal.

DOI： 10.1145/3427477.3429462

Language：English   Publishing type：Research paper (other academic)  

In this paper, we propose ProceThings, a new middleware platform to provide smart community services by utilizing computational resources of IoT devices in a target community area. To realize ProceThings, we address three key challenges: (1) dynamic load balance management among numerous IoT devices; (2) distributed task assignment/execution over the IoT devices taking into account fault-Tolerance; and (3) fulfillment of a service level agreement (SLA) for each service. For (1), ProceThings employs a heuristic monitoring mechanism which hierarchically aggregates load and resource conditions from all IoT devices in the target area (or belonging to a service). For (2), ProceThings employs a cluster-based architecture where proximity IoT devices are grouped into clusters with a fail over function, where they are allocated processing tasks from user queries. For (3), ProceThings employs demand-Aware in-situ resource provisioning which dynamically predicts and assigns a sufficient amount of computational resources within the area where the service is provided to meet the SLA while preventing over-provisioning of resources. We have implemented a prototype of ProceThings running on commodity small computers consisting of Raspberry Pis and Intel NUCs and confirmed that the above mechanisms can properly work satisfying the corresponding SLAs when running smart community services.

DOI： 10.1145/3427477.3429275

Language：Japanese  

How the information technologies change the occupation

DOI： 10.32191/jjos.14.003

Language：Japanese  

Language：Others  

Language：Others  

itocon: Bus Stop Congestion Visualizing System with Multiple Congestion Sensors

Language：English   Publishing type：Research paper (other academic)  

Our objective is to achieve a city where everyone can move safely and comfortably by developing and implementing ICT-based mobile support system at the actual transport hub. Our system uses cameras installed at the transport hub to detect people who have difficulty moving, and notifies this information to the transportation staff in real time to help them move more smoothly. This system makes it possible to aggregate and provide information on places that is useful for COVID-19 measures, such as measuring the congestion of places and the social distances of the people who gather there.

DOI： 10.1145/3384419.3430609

Language：English   Publishing type：Research paper (other academic)  

As society becomes increasingly interconnected, the need for sophisticated signal processing and data analysis techniques becomes increasingly apparent, particularly in the field of Intelligent Transportation Systems (ITS) where various sensing applications generate data at an exponential rate. In this paper, we put a forward a compressive sensing-based system to extract information from passing vehicle sounds sampled at sub-Nyquist rates for Acoustic Vehicle Detection and Identification (AVDI) applications. The obtained compressive measurements are used to detect and identify passing vehicles. Initial evaluation performed using data obtained from roads on a university campus presents an accuracy of 86.2 &#37; with a back-end ADC sample rate of 3 kHz.

DOI： 10.1109/VTC2020-Fall49728.2020.9348569

Language：English   Publishing type：Research paper (other academic)  

Due to the spread of COVID-19, we are desired to avoid crowded places including public transportation. Kyushu University has the largest campus in Japan, called "Ito campus", and the population there is about 20,000 in which 23&#37; of students and 46&#37; of staff use a bus for reaching the campus. The lectures in the first half of 2020 have been conducted online, but we plan to resume face-to-face lectures gradually. At that time, we expect the bus stops and buses to be crowded, especially during rush hour. In this paper, we introduce a system, called Itocon, to visualize the human congestion of bus stops around the campus. Itocon aggregates the sensing data from various sensors deployed around the target bus stops, and calculate and visualize the congestion degrees in real-time. Itocon is developed as a web application to avoid requesting the application install. We hope all the people who use a bus change their moving time based on the congestion information for avoiding human crowds. We explain the details and the future prospects of Itocon.

DOI： 10.1145/3384419.3430395

Language：English   Publishing type：Research paper (other academic)  

In this paper, the authors present SMArch, an inter-organizational software architecture for smart mobility, which can be defined as services coordinating multi-modal mobility. The architecture is extended from the Japanese system architecture for intelligent transportation systems (ITS), which is defined in 1999 and used for a reference giving a shared vision among stakeholders involved in ITS. Although the Japanese ITS architecture focuses road traffic only, SMArch focuses any mobility including not only road vehicles but also airplanes and ships. Moreover, SMArch introduces the service path concept, handles the itinerary as a series of mobility services and staying services such as lodging, dining and parking. Service coordinating agents present service path candidates satisfying the given traveler's requirements. The agents and other services interact to search for such service path candidates with the service path query language defined based on the service path concept and the information model defined in the SMArch.

DOI： 10.1145/3436829.3436868

Language：English   Publishing type：Research paper (other academic)  

Smart tourism is attracting attention of researchers in recent years. Its technologies can be used by tourists in order to obtain useful information during sightseeing with smart devices etc. To provide suitable and personalized tourism information according to the situation of tourists, understanding psychological status during sightseeing, especially, emotional status and satisfaction level, is important. We assume that the psychological status of tourists is appearing and represented through unconscious behaviors during sightseeing such as head/body movements and facial/vocal expressions, and have proposed methods to estimate emotion and satisfaction statuses by sensing and analyzing tourists' behaviors. Through in-the-wild experiments with 22 participants, we found that the difference in tourists' attributes might give effects for the estimation. In this paper, we have statistically analyzed those effects, focusing on tourists' nationality. As a result of the two-way ANOVA, we found the interaction effect (disordinal interaction) between tourists' nationality and estimation performance, the main effect in differences of features, and the main effect in differences of tourists' nationality. The results imply that we need to take tourists' nationality into account for building estimation models. Contribution: We have statistically analyzed the nationality effects on tourist emotion and satisfaction estimation, and confirmed significant differences in feature contributions for estimation models.

DOI： 10.1109/ICIEVicIVPR48672.2020.9306525

Language：English   Publishing type：Research paper (scientific journal)  

In the tourism sector, user-generated information and communication among tourists are perceived to be more effective and reliable contents. In addition, the collection of dynamic tourism information with high spatio-temporal resolution is required to provide comfortable tourism in response to the changing tourism style with the advancement of information technology. Participatory sensing, which can collect various types of information is a useful method by which to collect these contents. However, continuous participation of users is essential in participatory sensing, and it is one of the most important points to stimulate participation motivation. In the tourism situation, we also need to pay attention to the total tourist satisfaction of participants. In this paper, we adopt gamification, i.e., the implementation of game design elements in real-world contexts for non-gaming purposes, for participatory sensing as an incentive mechanism to motivate participants with active participation and collect the necessary information efficiently. Within the framework, where points are given when completing the requested sensing task (=mission), two sensing missions with different burdens; Area Mission and Check-in Mission, and three different types of rewarding mechanisms; Fixed, Variable and Dynamic Variable, are designed as a gamification mechanism. We implemented these elements in the proposed participatory sensing platform application and conducted an experimental case study with 33 participants at an actual tourist spot: Kyoto, Japan. Then, we investigate the effects on tourist behavior and satisfaction by analyzing collected sensor data, mission logs, and post-survey answers. As a result, we can conclude the following: (1) the tourist behavior is changed due to the proposed gamification design and the necessary information was collected efficiently; (2) the participants tend to prioritize Check-in Mission over the sightseeing, which can induce a behavior change but might impact sightseeing enjoyment.

DOI： 10.3390/smartcities3030037

Language：English   Publishing type：Research paper (scientific journal)  

In this paper, we design and develop a sensor-embedded office chair that can measure the posture of the office worker continuously without disturbing their job. In our system, eight accelerometers, that are attached at the back side of the fabric surface of the chair, are used for recognizing the posture. We propose three sitting posture recognition algorithms by considering the initial position of the chair and the difference of physique. Through the experiment with 28 participants, we confirm that our proposed chair can recognize the sitting posture by 75.4% (algorithm 1), 83.7% (algorithm 2), and 85.6% (algorithm 3) respectively.

DOI： 10.1587/transinf.2019EDP7226

Language：English   Publishing type：Research paper (other academic)  

In Japan, the demand for nursing homes is increasing due to the increasing number of elderly people. Therefore, the burden on care work for nursing staff becomes very large. In a nursing home staff spend a large amount of time in making care reports which reduce the time for doing the essential care for residents (i.e., elderly people). Thus, to reduce the work burden for staff, the system on which staff can create reports easily and instantly is needed. In our previous work, to reduce the burden of making care reports, we have propose a location-based nursing record system, called "QuickCareRecord," which can automatically switch information input screen and its items depending on location. However, this work conducted only a limited evaluation by emulated environment. In this paper, we actually implement the proposed system by using BLE beacon and small terminal based on our previous work. To show the effectiveness of the proposed system, we carry out the experiment in the actual nursing home. As a result, we show that the proposed system can identify the care location and care recipient approximately, staff can make care records easily and accurately.

DOI： 10.1109/ISMICT48699.2020.9152614

Language：English   Publishing type：Research paper (scientific journal)  

In this paper, we design and develop a sensor-embedded office chair that can measure the posture of the office worker continuously without disturbing their job. In our system, eight accelerometers, that are attached at the back side of the fabric surface of the chair, are used for recognizing the posture. We propose three sitting posture recognition algorithms by considering the initial position of the chair and the difference of physique. Through the experiment with 28 participants, we confirm that our proposed chair can recognize the sitting posture by 75.4% (algorithm 1), 83.7% (algorithm 2), and 85.6% (algorithm 3) respectively.

DOI： 10.1587/transinf.2019EDP7226

Language：Others   Publishing type：Research paper (other academic)  

Nowadays, lifestyle diseases have become a global social problem. As a treatment, it is necessary to review people’s life patterns and lead them to a better lifestyle, which is called “Behavior Change.” With the progress of information technology, many pieces of research and developments have been conducted to assist people with their daily life, trying to induce behavior change. The common approach used in the existing systems is altering users’ target behaviors to new behaviors, such as drinking more water and walking further. However, this approach considers less about the timing of proposing and the relevance between users’ target behaviors and new behaviors, which limits the effectiveness of the existing systems in inducing behavior change. To improve users’ experience and effectiveness of the system, in this paper, we propose a new design of behavior change support system which promotes behavior changes by embedding additional behaviors into users’ daily routines, trying to improve users’ awareness of their health condition.

Language：Others   Publishing type：Research paper (scientific journal)  

Recently, due to the drastic increase in foreign tourists coming to Japan, there has been a demand to provide smart tourism services that enable inbound tourists to comfortably enjoy sightseeing. To provide satisfactory experiences for tourists, it is desirable to provide tourist information in a timely manner by considering dynamic information, which is information that changes over time, such as current congestion information in destination spots and travel route information, in addition to static information, such as the preferences and profiles of tourists. However, in many existing systems, serious problems occur, such as (1) a lack of support for on-site use, (2) a lack of consideration of dynamic information, and (3) heavy burden on tourists. In this paper, we propose a novel system that can provide tourism plans for tourism spots in a timely manner. The proposed system consists of the following two key mechanisms: (A) A mechanism for acquiring preference information from tourists (including preference on dynamic information); (B) a curation mechanism for realizing on-site tourism. To demonstrate the effectiveness of the proposed system, we carried out evaluation experiments utilizing real tourism spots and simulations. As a result, we obtained the following primary findings: (1) On-site tourism spot recommendation is effective for tourists who do not make detailed tourism plans before sightseeing; (2) preference information for participants can be reflected in the tourism spot recommendation while massively reducing the burden on participants; (3) it is possible to obtain a higher satisfaction level than is achieved with model courses, which are often used for sightseeing.

DOI： 10.3390/smartcities3020013

Language：Others   Publishing type：Research paper (other academic)  

A key feature of Intelligent Transport Systems (ITS) is the ability to detect and identify vehicles. In this paper, we put forward a stereo microphone-based system capable of detecting and identifying the type of individually, sequentially, and simultaneously passing vehicles in multi-lane environments based on their sound. We find that our proposed system shows improved performance over single-microphone systems thanks to its improved sequential and successive vehicle detection performance. Initial evaluation results using sound data collected from roads on a university campus show a classification accuracy of 95.01 &#37;.

DOI： 10.1109/SAS48726.2020.9220076

Language：English   Publishing type：Research paper (other academic)  

In data collection of the human physiological and psychological conditions for mental healthcare (e.g., work engagement), measurement methods using environment-installed sensors and questionnaire surveys have been often used. However, these approaches are not practical in continuous data collection, due to the large burden for people. Recently, in association with advancing sensing technology with IoTs, sensing by small sensors and wearable devices has become possible easily. In this paper, we aim to establish a simple and general sensing method based on a mobile application for measuring physiological and psychological state of office workers and environmental state. Through the experiment for 2-3 weeks involving 60 office workers of four Japanese companies by using our application, we succeeded to create a dataset of physiological, environment, and mental state. This paper explains the developed mobile application, experimental procedure, and a summary of the data collected in the experiment.

DOI： 10.1109/PerComWorkshops48775.2020.9156085

Language：Japanese  

SMArch : Software Architecture for Smart Mobility (ITS)

Language：English   Publishing type：Research paper (other academic)  

In recent years, researches on recognizing daily behavior and psychological/physiological states have been actively conducted to change the behavior of workers working in companies. In this paper, we analyzed Occupational Health questionnaire named DAMS for waking-up time and daily sleep data that are acquired from wearable devices in 2-3 weeks experiment of 60 office workers working in five general companies. By using a machine learning method, our binary Balanced Random Forest model predicts depression, positive, and anxiety moods in two levels, high and low. As a result of Leave One Person Out cross validation, it was confirmed that our model estimated with the F1 values of depression mood: 0.776, positive mood: 0.610, anxiety mood: 0.756. Moreover, we evaluated the variance of the three estimations among subjects by referencing the box chart. As a result, it was confirmed that there is variance in estimation accuracy for each subject.

DOI： 10.1109/PerComWorkshops48775.2020.9156176

Language：English   Publishing type：Research paper (other academic)  

As the adoption of Intelligent Transport Systems (ITS) grows worldwide, so does the need for lost-cost, fast-deployment vehicle detection systems. SAVeD is a low-cost acoustic detection system developed by the authors which works by fitting a curve indicating vehicle passage to a sound map depicting the difference in arrival time of a passing vehicle's sound at two microphones installed on the roadside. This paper expands on the SAVeD method by proposing a Two-Stage Acoustic Vehicle Detection System for use in high-traffic environments, where multiple simultaneously and successively passing vehicles cause interference in the detection process. To solve this problem, the sound map fitting process is divided into two stages: the detection range is narrowed based on information estimated during the Pre-Fitting stage, and neighborhood point extraction is performed during the Post-Fitting stage to improve vehicle detection accuracy. Initial evaluation performed on a four-lane, two-direction road showed a vehicle detection F-measure of 0.63, a 12-point increase over SAVeD.

DOI： 10.1109/PerComWorkshops48775.2020.9156248

Language：English   Publishing type：Research paper (other academic)  

In this paper, we proposed and implemented an augmented workplace where humans and deployed sensors have some interaction. Through the experiment, CO2 sensor asks humans to open the window when CO2 level exceeds the threshold, we found that a human has obeyed a message from a sensor almost every time.

DOI： 10.1145/3384657.3385334

Language：English   Publishing type：Research paper (scientific journal)  

Smartphone-based disaster communication systems have been the topic of many recent studies. However, existing systems unrealistically assume that prior to a disaster, the required applications have been installed in the smartphones of all users. In disaster areas without Internet access, obtaining the software required for communications systems is difficult because application distribution platforms, such as Google Play, are web-based. In this paper, we present RecurShare, which is a novel application distribution system. The key idea of RecurShare is to be able to distribute applications without the Internet by running a web server that can be accessed using the default web browser and the tethering function of Android smartphones. We evaluated the feasibility of RecurShare experimentally and showed that RecurShare functions without problems on a variety of commercial devices. Also, user feedback suggests that RecurShare sufficiently guides users through the application distribution process in a manner that is easy to understand. Furthermore, we found that RecurShare can distribute itself at an exponential rate and that it can transmit files at speeds that are comparable to those of commonly used wireless communication systems.

DOI： 10.18494/SAM.2020.2614

Language：English   Publishing type：Research paper (scientific journal)  

In sensor networks, sensor localization is one of the mandatory functions for physical sensing. In this paper, we propose a sensor localization system that utilizes wireless local area network (WLAN) access points (APs) as anchor nodes. WLAN is widely available nowadays, and WLAN APs are used in many environments. Therefore, for our localization system, we utilize the WLAN APs already installed in the environment, thereby eliminating the installation of new anchors. We present an AP received signal strength (RSS) measurement method, which is the key technology of our sensor localization system. Notably, sensor nodes equipped with an IEEE 802.15.4 (ZigBee) module cannot detect WLAN AP signals. We therefore developed an AP signal detection method with AP identification by employing a cross-technology signal detection technique. Experimental evaluations demonstrate that our AP signal detection method successfully identified sender APs with a false positive (FP) detection rate less than 20%. The experiments also validated that the measured AP-RSS followed the same distribution as the RSS measured using a WLAN device.

DOI： 10.18494/SAM.2020.2592

Language：English   Publishing type：Research paper (scientific journal)  

In the field of sports, there arc increasing opportunities to use inertial measurement units (IMUs) to enhance the training process and improve the performance of athletes. We focus on kendo, a traditional martial art using shinai (bamboo swords) in Japan, and propose methods for detecting and recognizing strike activities using IMUs towards realizing a kendo skill improvement support system. We used a sensor data set of strike activities obtained from 14 participants (seven kendo-experienced and seven inexperienced persons). We attached four IMUs to the participants' right wrist, waist, and shinai (tsuba and saki-gawa). First, to detect the strike activity, we calculated the dynamic time warping (DTW) distance between the training data and the time series data, and detected the strike activity sections. The proposed method detected strike activities with a high accuracy of 95.0%. Next, to recognize the strike activity, we recognized five types (Center-Men, Right-Men, Left-Men, Do, and Kote). In the person-dependent (PD) case, we achieved an accuracy of 89.5% using data of the right wrist. In the person-independent (PI) case, we achieved an accuracy of 54.9% using IMUs attached to the three positions. These results clarified the points to be improved in the proposed method to realize the support system.

DOI： 10.18494/SAM.2020.2615

Language：English   Publishing type：Research paper (scientific journal)  

DOI： 10.1002/1348-9585.12132

Language：English   Publishing type：Research paper (scientific journal)  

Smartphone-based disaster communication systems have been the topic of many recent studies. However, existing systems unrealistically assume that prior to a disaster, the required applications have been installed in the smartphones of all users. In disaster areas without Internet access, obtaining the software required for communications systems is difficult because application distribution platforms, such as Google Play, are web-based. In this paper, we present RecurShare, which is a novel application distribution system. The key idea of RecurShare is to be able to distribute applications without the Internet by running a web server that can be accessed using the default web browser and the tethering function of Android smartphones. We evaluated the feasibility of RecurShare experimentally and showed that RecurShare functions without problems on a variety of commercial devices. Also, user feedback suggests that RecurShare sufficiently guides users through the application distribution process in a manner that is easy to understand. Furthermore, we found that RecurShare can distribute itself at an exponential rate and that it can transmit files at speeds that are comparable to those of commonly used wireless communication systems.

DOI： 10.18494/SAM.2020.2614

Language：English   Publishing type：Research paper (scientific journal)  

Effective Trilateration-based Indoor Localization Method Utilizing Active Control of Lighting Devices

Language：Others  

DOI： 10.1002/1348-9585.12132

Language：English   Publishing type：Research paper (scientific journal)  

In sensor networks, sensor localization is one of the mandatory functions for physical sensing. In this paper, we propose a sensor localization system that utilizes wireless local area network (WLAN) access points (APs) as anchor nodes. WLAN is widely available nowadays, and WLAN APs are used in many environments. Therefore, for our localization system, we utilize the WLAN APs already installed in the environment, thereby eliminating the installation of new anchors. We present an AP received signal strength (RSS) measurement method, which is the key technology of our sensor localization system. Notably, sensor nodes equipped with an IEEE 802.15.4 (ZigBee) module cannot detect WLAN AP signals. We therefore developed an AP signal detection method with AP identification by employing a cross-technology signal detection technique. Experimental evaluations demonstrate that our AP signal detection method successfully identified sender APs with a false positive (FP) detection rate less than 20%. The experiments also validated that the measured AP-RSS followed the same distribution as the RSS measured using a WLAN device.

DOI： 10.18494/SAM.2020.2592

Language：English   Publishing type：Research paper (scientific journal)  

In sensor networks, sensor localization is one of the mandatory functions for physical sensing. In this paper, we propose a sensor localization system that utilizes wireless local area network (WLAN) access points (APs) as anchor nodes. WLAN is widely available nowadays, and WLAN APs are used in many environments. Therefore, for our localization system, we utilize the WLAN APs already installed in the environment, thereby eliminating the installation of new anchors. We present an AP received signal strength (RSS) measurement method, which is the key technology of our sensor localization system. Notably, sensor nodes equipped with an IEEE 802.15.4 (ZigBee) module cannot detect WLAN AP signals. We therefore developed an AP signal detection method with AP identification by employing a cross-technology signal detection technique. Experimental evaluations demonstrate that our AP signal detection method successfully identified sender APs with a false positive (FP) detection rate less than 20%. The experiments also validated that the measured AP-RSS followed the same distribution as the RSS measured using a WLAN device.

DOI： 10.18494/SAM.2020.2592

Language：Japanese   Publishing type：Research paper (scientific journal)  

Language：Japanese  

Language：English   Publishing type：Research paper (other academic)  

Walking has been attracting attention as an important means for prevention and improvement of lifestyle diseases, such as high blood pressure and diabetes. However, walking at a continuous pace with high load can be challenging, and health benefits cannot be expected when walking with low load. A walking pace support system is necessary in order to achieve effective walking. Hence, we are developing BeatSync, a smartphone application that realizes the induction of a natural and accurate walking pace by the rhythm of music. This application can select songs from a user’s music library. However, some songs, such as songs with a fast (slow) rhythm or complex beats, are not suitable for walking pace induction. In the present paper, we consider the speed and clarity of music rhythm to be an important factor in selecting a song that is suitable for walking pace induction and make an index and clarify its effect on walking pace induction. In the present study, BPM is used as an index of rhythm speed, and RhythmValue (RV) is proposed as an index of rhythm clarity. In order to verify the effectiveness of the index, we conducted walking pace induction experiments using 30 songs with different speeds and clarities (two sets of 15 songs) with 14 participants. As a result, the experiments confirmed that the proposed index can distinguish songs that are suitable or unsuitable for walking pace induction and can select songs that are suitable for walking pace induction.

DOI： 10.1007/978-3-030-49289-2_5

Language：English   Publishing type：Research paper (other academic)  

By the spread of active learning and group work, the ability to collaborate and discuss among the participants becomes more important than before. Although several studies have reported on that micro facial expressions and body movements give psychological effects to others during conversation, most of them are lacking in quantitative evaluation and there are few datasets about group discussion. In this research, we proposed a highly reproducible system that helps to make datasets of group discussions with multiple devices such as an omnidirectional camera (360-degree camera), an eye tracker and a motion sensor. Our system operates those devices in one-stop to realizing synchronized recording. To confirm the feasibility, we built the proposed system with an omnidirectional camera, 4 eye trackers, and 4 motion sensors. Finally, we succeeded to make a dataset by recording 8 times group meeting by using our developed system easily.

Language：English   Publishing type：Research paper (scientific journal)  

Changing behavior related to improper lifestyle habits has attracted attention as a solution to prevent lifestyle diseases, such as diabetes, heart disease, arteriosclerosis, and stroke. To drive health behavior changes, wearable devices are needed, and they must not only provide accurate sensing and visualization functions but also effective intervention functions. In this paper, we propose a health support system, WaistonBelt X, that consists of a belt-type wearable device with sensing and intervention functions and a smartphone application. WaistonBelt X can automatically measure a waistline with a magnetometer that detects the movements of a blade installed in the buckle, and monitor the basic activities of daily living with inertial sensors. Furthermore, WaistonBelt X intervenes with the user to correct lifestyle habits by using a built-in vibrator. Through evaluation experiments, we confirmed that our proposed device achieves measurement of the circumference on the belt position (mean absolute error of 0.93 cm) and basic activity recognition (F1 score of 0.95) with high accuracy. In addition, we confirmed that the intervention via belt vibration effectively improves the sitting posture of the user.

DOI： 10.3390/s19204600

Language：Japanese  

A Solution of Vehicle Unevenness Problem by Potential Users toward a Realization of One-Way Car Sharing
The solution of the vehicle uneven distribution problem is an important issue in one-way car sharing. Although the existing method has tried to solve the problem by efficient vehicle movement by the operating company, there was a problem of cost. Therefore, as a new approach, we propose a method of requesting a potential user to move a vehicle. This proposal is set as a problem to maximize the number of required trips while reducing the number of request trips. As a result of evaluating by simulation to verify the effectiveness of the proposed method, even if the acceptance rate of the request trip was 20%, the acceptance rate of the request trip improved by 17%. In addition, in order to examine potential users to be requested, real data of Park 24 Co., Ltd. was analyzed. As a result, we found that there are 5 patterns (regular, 2 way, distributed, local, 1 route) as usage patterns. We also considered that usage patterns of users to be requested are distributed type and local type. The contribution of this research is to propose a new approach to solve the vehicle uneven distribution problem and to discuss not only simulation but also real data analysis.

Language：English   Publishing type：Research paper (scientific journal)  

Changing behavior related to improper lifestyle habits has attracted attention as a solution to prevent lifestyle diseases, such as diabetes, heart disease, arteriosclerosis, and stroke. To drive health behavior changes, wearable devices are needed, and they must not only provide accurate sensing and visualization functions but also effective intervention functions. In this paper, we propose a health support system, WaistonBelt X, that consists of a belt-type wearable device with sensing and intervention functions and a smartphone application. WaistonBelt X can automatically measure a waistline with a magnetometer that detects the movements of a blade installed in the buckle, and monitor the basic activities of daily living with inertial sensors. Furthermore, WaistonBelt X intervenes with the user to correct lifestyle habits by using a built-in vibrator. Through evaluation experiments, we confirmed that our proposed device achieves measurement of the circumference on the belt position (mean absolute error of 0.93 cm) and basic activity recognition (F1 score of 0.95) with high accuracy. In addition, we confirmed that the intervention via belt vibration effectively improves the sitting posture of the user.

DOI： 10.3390/s19204600

Language：English   Publishing type：Research paper (other academic)  

We propose a room-level place recognition system which uses energy harvesting elements as both sensor and power source. As energy harvesting elements, we focus on solar cells. In our previous work, by combining multiple types of cells and learning the difference of the generated power, we succeeded to distinguish ten places thanks to different characteristics of solar cells according to their materials. However, since the previous research focused on investigating the feasibility of our fundamental idea, it used a microcomputer board that requires an additional battery to measure the amount of generated power. Hence, it did not reach the level to utilize the energy harvesting elements as a power supply required for running the whole system. In this paper, we re-design our previously proposed mechanism to measure the amount of generated power of the energy harvesting elements and propose the novel mechanism to measure the amount of generated power by using the generated power without any additional battery. In stead of measuring the amount of generated power, our system has the circuit that counts how many times the electrolytic capacitor has been fully charged. The capacitor used is the minimum size required for reading the time stamp and write it to the memory. Finally, we can measure the amount of generated power by counting the time stamps recorded per unit of time. We designed and implemented the new circuit on PCB, and applied our system to room-level place recognition. As a result, we confirmed that our system can accurately distinguish the eight places.

DOI： 10.1109/SENSORS43011.2019.8956580

Language：Others   Publishing type：Research paper (other academic)  

This paper is the first trial to create a corpus on human-to-human multi-modal communication among multiple persons in group discussions. Our corpus includes not only video conversations but also the head movement and eye gaze. In addition, it includes detailed labels about the behaviors appeared in the discussion. Since we focused on the micro-behavior, we classified the general behavior into more detailed behaviors based on those meaning. For example, we have four types of smile: response, agree, interesting, sympathy. Because it takes much effort to create such corpus having multiple sensor data and detailed labels, it seems that no one has created it. In this work, we first attempted to create a corpus called “M3B Corpus (Multi-Modal Meeting Behavior Corpus),” which includes 320 minutes discussion among 21 Japanese students in total by developing the recording system that can handle multiple sensors and 360-degree camera simultaneously and synchronously. In this paper, we introduce our developed recording system and report the detail of M3B Corpus.

DOI： 10.1145/3341162.3345588

Language：Japanese  

Considerations on affectors of reliability for answer to online questionnaires : From point of style of questionnaire and context of smart phone manipulation

Language：Others   Publishing type：Research paper (scientific journal)  

DOI： 10.2197/ipsjjip.27.614

Language：English   Publishing type：Research paper (scientific journal)  

In recent years, many types of research and developments on behavior change have been conducted. The purpose of behavior change is to improve people's lifestyle pattern or to maintain the improvement for a long time with the aim to achieve a goal such as promoting health condition improvement. To achieve the foundation of a new lifestyle, it is necessary to recognize the daily life patterns of users and give triggers for behavior change to users in their daily life. To realize this, in our research, we develop an interactive signage which is able to identify and actively talk to the passing user and try to induce behavior change by sending visual and auditory stimulation. Then, we record users' reactions and upload them to the server. In this paper, we report the investigation result on users' reactions and feelings to the developed interactive signage. As a survey experiment, we set up four interactive signs on a floor of our university and asked 15 participants to carry a name tag with a Bluetooth Low Energy beacon during their daily life. Five kinds of tasks based on dialogue scenarios are posted to the approaching participants. Participants can respond to these tasks through a touchscreen. The period of the experiment was three weeks. To get the data in an ideal environment, during the first week, we asked all the participants to respond to the utterance from the interactive signage whenever they hear the voice message. During the next two weeks, participants were not asked to respond to the task definitely to get the data in the real environment. The result of the experiment showed that our proposed interactive signage could induce behavior change effectively. Based on the result of experiment, we updated our interactive signage system by adding response time (the time passed from showing contents until user respond), record function, and voice feedback function. Furthermore, to collect the data of response time that is considered as a part of users' reactions, we conducted an additional experiment with the same participants in previous experiment (except for one missing participant) for one week after updating the system. In the additional experiment, the participants were not asked to respond to the utterance definitely. As a result, it is shown that the behavior change by the proposed signage is still effectively induced. We also analyzed the relationship between the day passed and the response rate of each task type. The result shows that the number of ignorance of personal task and check task does not rise even as the time passes.

DOI： 10.1111/coin.12234

Language：English   Publishing type：Research paper (other academic)  

DOI： 10.1007/978-3-030-22015-0_29

Language：Japanese  

Language：Japanese  

Language：English   Publishing type：Research paper (scientific journal)  

DOI： 10.1504/IJAHUC.2019.100779

Language：Japanese  

Language：Japanese  

Language：English   Publishing type：Research paper (scientific journal)  

DOI： 10.1155/2019/9320145

Language：Japanese   Publishing type：Research paper (scientific journal)  

Language：English   Publishing type：Research paper (other academic)  

In Japan, the demand for nursing home is increasing due to the increase in the number of elderly people. However, the work burden for nursing staffs is very large. Specially, they spend much time in making care reports which reduces the time for doing the essential care for residents in a nursing home. Therefore, to reduce the load for staffs, the system on which staffs can create reports easily and instantly is needed. In this paper, we propose a mobile memo system that switches the view according to the location automatically and assists inputting the necessary information. Our proposed system enables care staffs to prepare the care report about caring for the elderly person quickly. Through the evaluation in the emulated environment, we show that our system succeed to reduce the time for recording by 37.9% and 58.9%, compared with the conventional system and traditional handwriting, respectively.

DOI： 10.1109/ISMICT.2019.8743900

Language：English   Publishing type：Research paper (other academic)  

In this paper, we propose a novel time-lapse based video summarization method without substantial loss of information important for viewers to plan a tour route. In the proposed method, we focus on scene changes in the video. We extract scenes with a certain level of changes compared to previous scenes as important (slowly played back) in the summarized video, while other scenes are fast forwarded. To evaluate the effectiveness of our proposed method, comparing the video summarized with our method and that summarized manually (by voting for necessary/unnecessary scenes), our method identified the important scenes with an F-measure of 62.22%.

DOI： 10.1109/MIPR.2019.00079

Language：English   Publishing type：Research paper (other academic)  

Smart mobility systems, which include Intelligent Transportation Systems(ITS) and smart energy systems, become more important. There is, however, lack of its platform studies. We started a sustainable information infrastructure project for smart mobility systems. The project pursues issues that establish an information infrastructure architecture and seamless development method chain for it. The project has mainly two features; 1) applying life-cycle-oriented methods, which are a cycle from system development to operations, to a real world, and 2) dealing with an uncertainty in system design phase. This paper describes current status of the project.

DOI： 10.1109/IIAI-AAI.2018.00025

Language：English   Publishing type：Research paper (scientific journal)  

DOI： 10.1109/PERCOMW.2019.8730659

Language：Others   Publishing type：Research paper (other academic)  

Following the technical progress and growing touristic market, demand on guidance systems is constantly increasing. Current systems are not personalized, they usually provide only a general information on sightseeing spot and do not concern about the tourist's perception of it. To design more adjustable and context-aware system, we focus on collecting and estimating emotions and satisfaction level, those tourists experience during the sightseeing tour. We reducing changes in their behaviour by collecting two types of information: conscious (short videos with impressions) and unconscious (behavioural pattern recorded with wearable devices) continuously during the whole tour. We have conducted experiments and collected initial data to build the prototype system. For each sight of the tour, participants provided an emotion and satisfaction labels. We use them to train unimodal neural network based models, fuse them together and get the final prediction for each recording. As tourist himself is the only source of labels for such system, we introduce an approach of post-experimental label correction, based on paired comparison. Such system built together allows us to use different modalities or their combination to perform real-time tourist emotion recognition and satisfaction estimation in-the-wild, bringing touristic guidance systems to the new level.

DOI： 10.1109/PERCOMW.2019.8730864

Language：English   Publishing type：Research paper (other academic)  

In this paper, we focus on Kendo, which is a traditional sport in Japan, and propose a strikes-thrusts activity recognition method using a wrist sensor towards a pervasive Kendo support system. We collected the inertial sensor data set from 6 subjects. We attached 3 inertial sensor units (IMUs) on the subjects body, and 2 IMUs on the Shinai (bamboo sword used for Kendo). On the body, IMUs were placed on the Right Wrist, Waist and Right Ankle. On the Shinai, they were placed on the Tsuba and Saki-Gawa. We first classified strikes-thrusts activities consisting of 4 general types, Men, Tsuki, Do, and Kote, followed by further classification into 8 detailed types. We achieved 90.0&#37; of F-measure in the case of 4-type classification and 82.6&#37; of F-measure in the case of 8-type classification when learning and testing the same subjects data for only Right Wrist. Further, when adding data of sensors attached to the Waist and Right Ankle, we achieved 97.5&#37; of F-measure for 4-type classification and 91.4&#37; of F-measure for 8-type classification. As a result of leave-one-person-out cross-validation from 6 subjects to confirm generalized performance, in the case of 4-type classification, we achieved 77.5&#37; of F-measure by using only 2 IMUs (Right Wrist and Shinai Tsuba).

DOI： 10.1109/PERCOMW.2019.8730861

Language：English   Publishing type：Research paper (other academic)  

Recently, various smart home services such as smart air-conditioning, monitoring of elderly/kids and energy-efficient appliance operations are emerging, thanks to technologies of indoor positioning of users and recognition of Activity of Daily Living (ADL). Meanwhile, to realize more convenient home services, it will become more important to be able to predict occurrence time of each ADL. ADL prediction is a challenging problem since it is difficult to train a prediction model by general machine learning algorithms which use only the data at a moment for classification. In this paper, taking into account temporal dependency of data (consumed power of appliances and position of users) collected during daily life, we propose a method for constructing models to predict ADL with LSTM (Long Short-Term Memory). In the proposed method, we construct LSTM-based models by setting occurrence time of each activity to an objective variable. First, we tried to construct a multi-class classification model which outputs one of several predefined time ranges (time elapsed from present) as the occurrence time of the activity. Through preliminary experiment, we found that this model results in low accuracy in predicting the occurrence time. Then, as the second approach, we constructed a before-or-after classification model which judges if the activity occurs within a specified time or not. We applied this model to our smart home data and confirmed that it achieves better prediction accuracy for all activities.

DOI： 10.1109/PERCOMW.2019.8730662

Language：Others   Publishing type：Research paper (other academic)  

DOI： 10.1109/PERCOMW.2019.8730712

Language：English   Publishing type：Research paper (scientific journal)  

A wearable based long-term lifelogging system is desirable for the purpose of reviewing and improving users lifestyle habits. Energy harvesting (EH) is a promising means for realizing sustainable lifelogging. However, present EH technologies suffer from instability of the generated electricity caused by changes of environment, e.g., the output of a solar cell varies based on its material, light intensity, and light wavelength. In this paper, we leverage this instability of EH technologies for other purposes, in addition to its use as an energy source. Specifically, we propose to determine the variation of generated electricity as a sensor for recognizing "places" where the user visits, which is important information in the lifelogging system. First, we investigate the amount of generated electricity of selected energy harvesting elements in various environments. Second, we design a system called EHAAS (Energy Harvesters As A Sensor) where energy harvesting elements are used as a sensor. With EHAAS, we propose a place recognition method based on machine-learning and implement a prototype wearable system. Our prototype evaluation confirms that EHAAS achieves a place recognition accuracy of 88.5% F-value for nine different indoor and outdoor places. This result is better than the results of existing sensors (3-axis accelerometer and brightness). We also clarify that only two types of solar cells are required for recognizing a place with 86.2% accuracy.

DOI： 10.1109/PERCOM.2019.8767385

Language：English   Publishing type：Research paper (other academic)  

Recently, smart speakers like Amazon Echo and Google Home have been spread widely. Those devices support users' life through voice interface by receiving voice commands to operate appliances and order goods to online shops. Meanwhile, it is reported that smart speakers are vulnerable to some malicious attacks which steal personal information and/or order unnecessary goods by uttering voice from a device nearby the speaker, abusing the fact that the smart speakers cannot distinguish human voice from machine voice. A new type of attack called DolphinAttack which utters ultrasonic voice inaudible to human is also reported. Therefore, a method to identify which of human or machine is sending voice commands to a smart speaker is desired. In this paper, to prevent such machine-voice based attacks to a smart speaker in absence of residents, we propose a system consisting of a speaker and a microphone array to detect the existence of a human nearby, supposing it can be incorporated in a smart speaker in the future. In our proposed system, the speaker emits sonar sound generated based on Orthogonal Frequency Division Multiplexing (OFDM) in all directions, the microphone array with 8 channels attached on top of the speaker receives the reflected sound, and the human existence is judged by comparing the reflected sound with that measured in the same environment without human. Through experiments with a prototype system, we confirmed that our proposed system can detect the human existence by measuring the reflected signal of 0.5 second.

DOI： 10.1109/PERCOMW.2019.8730674

Language：English   Publishing type：Research paper (other academic)  

In this paper, we construct an interactive signage system, which is able to actively talk to users who are passing the signage system and at the same time tries to induce behavioral changes through visual and auditory stimulation. On top of our previous feasibility study, we re-design our proposed system based on the persuasive system design (PSD) model and the behavior change support system (BCSS) theory. The new design consists of 4 parts: recognizing (identifying and classifying users), executing (sending triggers and inducing behavior change), reviewing (recording users' reaction), and feedback (keeping or improving users' motivation). To track users' status and performance, we use the smartphone and the smartwatch to collect users' bio-data (e.g. heart rate, number of steps) and users' location. To send user triggers continuously without being interrupted by notifications from other applications, we set the interactive signage in the daily action line of users and present information to users when they pass the signage. In order to make the system more persuasive, we pick up eight features from the 28-feature list of PSD and apply them to our design.

DOI： 10.1109/PERCOMW.2019.8730807

Language：Others   Publishing type：Research paper (other academic)  

Many interrupt notification methods have been studied, but most of the existing research assumes that, except for the target application, the applications do not control the notification timing. However, if other applications are controlled by the same notification timing, interrupt timing will become concentrated and the effects of notification timing control may not be exerted. In addition, since the installed applications are different for each user, it is necessary to control notification timings while taking into consideration the behaviors of all the applications installed on a user's smartphone. In this research, we define notification timing control while considering the behaviors of all installed applications as 'Adaptive Notification Management' and then conduct diversity surveys of notifications received by users. In this paper, we report on a system that acquires all notification information while excluding privacy data. We also report on experimental results using actual data collected using crowdsourcing and discuss how to realize the application realizing adaptive notification management.

DOI： 10.1109/PERCOMW.2019.8730694

Language：English   Publishing type：Research paper (other academic)  

Walking has been attracting attention as an important means for prevention and improvement of lifestyle diseases such as high blood pressure and diabetes. In contrast of its importance, not many people know the appropriate pace or amount of walking for themselves. In our previous research, we proposed a system to recommend a route with appropriate walking speed/pace for consuming target calories while keeping load within a moderate range by predicting heart rate variation if walking along the road. However, no system is available which enables a user to naturally adjust his/her walking speed to the specified one while walking. In this paper, focusing on the entertainment effect of the music, we propose a smartphone application called BeatSync which leads everyone to walk naturally and accurately toward the target walking speed.

DOI： 10.1109/PERCOMW.2019.8730833

Language：English   Publishing type：Research paper (other academic)  

In this paper, we propose a novel time-lapse based video summarization method without substantial loss of information important for viewers to plan a tour route. In the proposed method, we focus on scene changes in the video. We extract scenes with a certain level of changes compared to previous scenes as important (slowly played back) in the summarized video, while other scenes are fast forwarded. To evaluate the effectiveness of our proposed method, comparing the video summarized with our method and that summarized manually (by voting for necessary/unnecessary scenes), our method identified the important scenes with an F-measure of 62.22&#37;.

DOI： 10.1109/MIPR.2019.00079

Language：Japanese  

Language：Others   Publishing type：Research paper (other academic)  

DOI： 10.1109/PERCOMW.2019.8730659

Language：Japanese   Publishing type：Research paper (scientific journal)  

Language：English   Publishing type：Research paper (other academic)  

In recent years, many research and developments of behavior change have been conducted. The purpose of behavior change is to improve people's lifestyle pattern or to maintain the improvement for a long time aiming to achieve a goal of promoting health condition improvement. In order to achieve the foundation of a new lifestyle, it is necessary to recognize the daily life patterns of users and give triggers to users in their daily life for behavior change. To realize this, in this study, we construct an interactive signage which is able to positively talk to the passing user and try to induce behavior change by sending visual stimulation and auditory stimulation. In this paper, we report the result of the investigation on users' reactions and feelings to the proposed interactive signage. As a survey experiment, we set up 4 interactive signage on a floor of our university and asked 15 participants to carry a name tag with a BLE beacon during their daily life. There are 4 different dialogue scenarios and a scenario is posted to the participants when he/she approaches the interactive signage. Participants can respond to these scenarios through a touchscreen. Based on the result of the experiment, it is found that the interactive signage can induce behavior change effectively.

DOI： 10.23919/ICMU.2018.8653587

Language：Japanese   Publishing type：Research paper (scientific journal)  

Language：Japanese  

Evacuation Center Determination Method Considering Bias of Congestion Degree on Areas
The quick and appropriate evacuation of the disaster victims is one of great concern in the disaster situation. Under the disaster situations, it is general that the victims evacuate from the present location to the nearest evacuation center depending on evacuation guidance. However, in case of the situation that many victims are at biased distribution to specific areas, the guidance to the closest evacuation center concentrates many victims to particular evacuation. As a result, the victims who could not enter the closest evacuation center need to evacuate to other evacuation centers, until accepted. Therefore, the problem occurs that the evacuation of victims is delayed significantly. In this paper, to solve this problem, we propose an evacuation center determination method aimed at reducing the evacuation time of the whole victims. Specifically, the proposed method determines an evacuation center so that the evacuation time of all victims becomes short based on the location information of victims and the capacity of evacuation centers. To evaluate the effectiveness of the proposed method, we compared our proposed method with the shortest route selection method and random selection method, by using the simulation study assuming the situation of the Gion festival in Kyoto. As a result, when using the proposed method, we showed that the average evacuation time can be reduced to 35.9% compared with the shortest route selection method. Also, we confirmed that the average evacuation time can be reduced to 8.1% even if victims who follow the proposed method is 20% of the total. From these results, even if victims who follow the proposed method is a part of the total, we showed that our method can reduce the overall evacuation time, and contribute to disaster prevention.

Language：Japanese  

A System for Searching MyMaps by Complementing Additional Information of POIs
In Google Maps and Open Street Map (OSM), a function which allows users to create own map including necessary Point of Interests (POIs) is provided. This function is called My Maps in Google Maps and uMap in OSM. A variety of maps have been created so far including maps that gather niche information. However, it is difficult to search My Maps that meets the purpose using keywords because these maps often do not contain sufficient information. In this paper, to improve the accuracy and intuitiveness in searching My Maps, we propose a novel system for searching MyMaps where a category ratio is used as a feature of MyMaps. To realize the system, we propose methods to complement the insufficient information composed of 1) identifying a probable POI in POI database with incomplete POI information and 2) estimating a category name of POI from POI name. We collected 46,996 POIs included in 5,526 public MyMaps for evaluation. As a result, we achieved 58.82&#37; accuracy in POI identification and 70.3&#37; accuracy in POI category estimation. We also implemented the prototype system for evaluating the convenience of the search by ten subjects. The result shows our proposed system can retrieve MyMaps including desired POI sets by using a category ratio as a feature.

Language：English   Publishing type：Research paper (other academic)  

In an era where billions of IoT devices are deployed, edge/fog computing paradigms are attracting attention for their ability to reduce processing delays and mitigate waste of communication resources. However, since the computing system assumed by edge/fog paradigms have heterogeneity (in terms of the computing power of devices, network performance between devices, device density, etc.), provisioning computational resources according to computational demand becomes a challenging constrained optimization problem. In this paper, we propose in-situ resource provisioning method consisting of insitu resource area selection with adaptive scale out and in-situ task scheduling based on tabu search algorithm. We conducted a simulation study in a target regional area where 2,000 IoT devices and 10 IoT services are deployed to evaluate the effectiveness of the proposed algorithm. The simulation results show that our proposed algorithm can obtain higher user QoS compared to conventional resource provisioning algorithms.

DOI： 10.1109/IWQoS.2018.8624127

Language：Others  

Improvement of Rent Estimation of Real Estate for Restaurants Using Catchphrase

Language：Japanese  

Design of Mobile Application with View Switching and Information Supplement Functions Depending on Current Positions of Users

Language：English   Publishing type：Research paper (scientific journal)  

Recently, companies have begun to care more about the well-being of their employees. With the spread of sensors, the Internet of Things, and artificial intelligence, the movement to build a better working environment by utilizing these technologies has been spreading. Especially, research on behavior that can change lifestyle habits is becoming popular. In this paper, we summarize workplace behavior research and projects for sensing and changing human behavior in a workplace and aim to improve the productivity and wellness of employees. Also, we introduce concepts for future workplaces and some of our related achievements. For physical state sensing, we have developed a continuous posture-sensing chair, which will soon be available commercially. For internal state sensing, we propose a method for estimating quality of life with wearable sensors. Our system have already achieved to estimate QoL (Quality of Life) around 90% with only 9 questions. In addition, we propose interactive digital signage to provide habit-changing reminders. Through one month experiment, we confirmed that our system can be feasible in daily life.

DOI： 10.2197/ipsjjip.27.614

Language：English   Publishing type：Research paper (scientific journal)  

Obtaining accurate floor plans of buildings is critical for optimising indoor geographic information system (GIS) applications. In this paper, we present a room measurement tool that utilises a smartphone equipped with an ultrasonic sensor. To take measurements, users complete a lap along the walls of all of the rooms. Then the tool accurately estimates the shape and size of them by tracking the walking paths of users and measuring the distance from the path to the walls with ultrasonic sensors. To track walking paths, we utilise inertial sensors embedded in the smartphone to estimate walking steps and turns, and the ultrasonic sensors to estimate the stride length when walking toward the wall. To account for such adjacent objects as bookshelves that decrease the accuracy of room size estimation, we used a mixed Gaussian filter. Our experimental results show that our tool considerably improved the estimation accuracy of the room shape and size.

DOI： 10.1504/IJAHUC.2019.100779

Language：English   Publishing type：Research paper (scientific journal)  

In recent years, many types of research and developments on behavior change have been conducted. The purpose of behavior change is to improve people's lifestyle pattern or to maintain the improvement for a long time with the aim to achieve a goal such as promoting health condition improvement. To achieve the foundation of a new lifestyle, it is necessary to recognize the daily life patterns of users and give triggers for behavior change to users in their daily life. To realize this, in our research, we develop an interactive signage which is able to identify and actively talk to the passing user and try to induce behavior change by sending visual and auditory stimulation. Then, we record users' reactions and upload them to the server. In this paper, we report the investigation result on users' reactions and feelings to the developed interactive signage. As a survey experiment, we set up four interactive signs on a floor of our university and asked 15 participants to carry a name tag with a Bluetooth Low Energy beacon during their daily life. Five kinds of tasks based on dialogue scenarios are posted to the approaching participants. Participants can respond to these tasks through a touchscreen. The period of the experiment was three weeks. To get the data in an ideal environment, during the first week, we asked all the participants to respond to the utterance from the interactive signage whenever they hear the voice message. During the next two weeks, participants were not asked to respond to the task definitely to get the data in the real environment. The result of the experiment showed that our proposed interactive signage could induce behavior change effectively. Based on the result of experiment, we updated our interactive signage system by adding response time (the time passed from showing contents until user respond), record function, and voice feedback function. Furthermore, to collect the data of response time that is considered as a part of users' reactions, we conducted an additional experiment with the same participants in previous experiment (except for one missing participant) for one week after updating the system. In the additional experiment, the participants were not asked to respond to the utterance definitely. As a result, it is shown that the behavior change by the proposed signage is still effectively induced. We also analyzed the relationship between the day passed and the response rate of each task type. The result shows that the number of ignorance of personal task and check task does not rise even as the time passes.

DOI： 10.1111/coin.12234

Language：English   Publishing type：Research paper (scientific journal)  

Medical studies have intensively demonstrated that sports activity can enhance both the mental and the physical health of practitioners. In recent years, fitness activity became the most common way to motivate and engage people in sports activity. Recently, there have been multiple attempts to elaborate on the “ideal” IoT-based solution to track and assess these fitness activities. Most fitness activities (except aerobic activities like running) involve one or multiple interactions between the athlete's hand palms and body or between the hand palms and the workout materials. In this work, we present our idea to exploit these biomechanical interactions of the hand palms to track fitness activities via a smart glove. Our smart glove-based system integrates force-sensitive resistor (FSR) sensors into wearable fitness gloves to identify and count fitness activity, by analyzing the time series of the pressure distribution in the hand palms observed during fitness sessions. To assess the performance of our proposed system, we conducted an experimental study with 10 participants over 10 common fitness activities. For the user-dependent activity recognition case, the experimental results showed 88.90% of the F score for overall activity recognition. The result of leave-one-participant-out cross-validation showed an F score ranging from 58.30% to 100%, with an average of 82.00%. For the exercise repetition count, the system achieved an average counting error of 9.85%, with a standard deviation of 1.38.

DOI： 10.1155/2019/9320145

Language：English   Publishing type：Research paper (other academic)  

In an era where billions of IoT devices have been deployed, edge/fog computing paradigms are attracting attention for their ability to reduce processing delays and communication overhead. In order to improve Quality of Experience (QoE) of regional IoT services that create timely geo-spatial information in response to users’ queries, it is important to efficiently allocate sufficient resources based on the computational demand of each service. However since edge/fog devices are assumed to be heterogeneous (in terms of their computational power, network performance to other devices, deployment density, etc.), provisioning computational resources according to computational demand becomes a challenging constrained optimization problem. In this paper, we formulate a delay constrained regional IoT service provisioning (dcRISP) problem. dcRISP assigns computational resources of devices based on the demand of the regional IoT services in order to maximize users’ QoE. We also present dcRISP+, an extension of dcRISP, that enables resource selection to extend beyond the initial area in order to satisfy increasing computational demands. We propose a provisioning algorithm, in-situ resource area selection with adaptive scale out and in-situ task scheduling based on a tabu search, to solve the dcRISP+ problem. We conducted a simulation study of a tourist area in Kyoto where 4,000 IoT devices and 3 types of IoT services were deployed. Results show that our proposed algorithms can obtain higher user QoE compared to conventional resource provisioning algorithms.

DOI： 10.1109/SEC.2018.00022

Language：Japanese  

Development of Summarization System for Electronic Health Record

DOI： 10.11517/jsaisigtwo.2018.aimed-006_01

Language：English   Publishing type：Research paper (scientific journal)  

With the spread of smart devices, people may obtain a variety of information on their surrounding environment thanks to sensing technologies. To design more context-aware systems, psychological user context (e.g., emotional status) is a substantial factor for providing useful information in an appropriate timing. As a typical use case that has a high demand for context awareness but is not tackled widely yet, we focus on the tourism domain. In this study, we aim to estimate the emotional status and satisfaction level of tourists during sightseeing by using unconscious and natural tourist actions. As tourist actions, behavioral cues (eye and head/body movement) and audiovisual data (facial/vocal expressions) were collected during sightseeing using an eye-gaze tracker, physical-activity sensors, and a smartphone. Then, we derived high-level features, e.g., head tilt and footsteps, from behavioral cues. We also used existing databases of emotionally rich interactions to train emotion-recognition models and apply them in a cross-corpus fashion to generate emotional-state prediction for the audiovisual data. Finally, the features from several modalities are fused to estimate the emotion of tourists during sightseeing. To evaluate our system, we conducted experiments with 22 tourists in two different touristic areas located in Germany and Japan. As a result, we confirmed the feasibility of estimating both the emotional status and satisfaction level of tourists. In addition, we found that effective features used for emotion and satisfaction estimation are different among tourists with different cultural backgrounds.

DOI： 10.3390/s18113978

Language：English   Publishing type：Research paper (scientific journal)  

With the spread of smart devices, people may obtain a variety of information on their surrounding environment thanks to sensing technologies. To design more context-aware systems, psychological user context (e.g., emotional status) is a substantial factor for providing useful information in an appropriate timing. As a typical use case that has a high demand for context awareness but is not tackled widely yet, we focus on the tourism domain. In this study, we aim to estimate the emotional status and satisfaction level of tourists during sightseeing by using unconscious and natural tourist actions. As tourist actions, behavioral cues (eye and head/body movement) and audiovisual data (facial/vocal expressions) were collected during sightseeing using an eye-gaze tracker, physical-activity sensors, and a smartphone. Then, we derived high-level features, e.g., head tilt and footsteps, from behavioral cues. We also used existing databases of emotionally rich interactions to train emotion-recognition models and apply them in a cross-corpus fashion to generate emotional-state prediction for the audiovisual data. Finally, the features from several modalities are fused to estimate the emotion of tourists during sightseeing. To evaluate our system, we conducted experiments with 22 tourists in two different touristic areas located in Germany and Japan. As a result, we confirmed the feasibility of estimating both the emotional status and satisfaction level of tourists. In addition, we found that effective features used for emotion and satisfaction estimation are different among tourists with different cultural backgrounds.

DOI： 10.3390/s18113978

Language：Japanese   Publishing type：Research paper (scientific journal)  

Language：English   Publishing type：Research paper (other academic)  

As the world's population of senior citizens continues to grow, the burdens on the professionals who care for them (carers) are also increasing. In nursing homes, carers need to make a daily report for each resident aiming to improve his/her quality of life. However, in the present understaffed situation, it is difficult and burdensome for carers to record the resident's activity in detail since each carer needs to take care of several residents at the same time. In this paper, we propose an automatic daily report generation system which can monitor the activity of multiple residents in nursing homes. Knowing that important activities such as toilet, bathing, rehabilitation and so on take place in specific areas in a nursing home, it is possible to record residents' activities by tracking their stay areas and movement between the areas within the day. Our proposed system estimates stay areas of multiple residents by machine learning for RSSI values that are sent from BLE beacons attached to residents and received at BLE scanners deployed over multiple areas, and records activities of the residents determined based on their estimated stay areas. The proposed system can also output a daily report of each resident based on the recorded data. We carried out a five-day experiment with four elderly participants in a nursing home and evaluated activity estimation accuracy by leave-one-person-out cross-validation. As a result, our proposed system achieved the weighted average F-measure of 81.6%.

DOI： 10.1109/PERCOMW.2018.8480348

Language：Others   Publishing type：Research paper (other academic)  

© Copyright held by the owner/author(s). To design more context-aware systems for smart environments, especially smart cities, the psychological user status such as emotion should be considered in addition to environmental information. In this study, we focus on the tourism domain as a typical use case, and propose a multimodal tourist emotion recognition method during the sightseeing. We employ behavioural cues (eye and head/body movements) and audio-visual features to recognise emotion. As a result of real-world experiments with tourists, we achieved up to 0.71 of average recall score in 3-class emotion recognition task with feature level fusion.

DOI： 10.1145/3267305.3267687

Language：English   Publishing type：Research paper (other academic)  

© 2018 Association for Computing Machinery In this paper, we, Ubi-NUTS Japan, introduce a multi-stage activity inference method that can recognize a user's mode of locomotion and transportation using mobile device sensors. We use the Sussex-Huawei Locomotion-Transportation (SHL) dataset to tackle the SHL recognition challenge, where the goal is to recognize 8 modes of locomotion and transportation (still, walk, run, bike, car, bus, train, and subway) activities from the inertial sensor data of a smartphone. We adopt a multi-stage approach where the 8 class classification problem is divided into multiple sub-problems considering the similarity of each activity. Multimodal sensor data collected from a mobile phone are inferred using a proposed pipeline that combines feature extraction and 4 different types of classifiers generated using the random forest algorithm. We evaluated our method using data from over 271 hours of daily activities of 1 participant and the 5-fold cross-validation. Evaluation results demonstrate that our method clearly recognizes the 8 types of activities with an average F1-score of 97&#37;.

DOI： 10.1145/3267305.3267526

Language：English   Publishing type：Research paper (other academic)  

DOI： 10.1145/3267305.3267504

Language：English   Publishing type：Research paper (other academic)  

In recent years, many research and developments of behavior change have been conducted. The purpose of behavior change is to improve people's lifestyle pattern or to maintain the improvement for a long time aiming to achieve a goal of promoting health condition improvement. In order to achieve the foundation of a new lifestyle, it is necessary to recognize the daily life patterns of users and give triggers to users in their daily life for behavior change. To realize this, in this study, we construct an interactive signage which is able to positively talk to the passing user and try to induce behavior change by sending visual stimulation and auditory stimulation. In this paper, we report the result of the investigation on users' reactions and feelings to the proposed interactive signage. As a survey experiment, we set up 4 interactive signage on a floor of our university and asked 15 participants to carry a name tag with a BLE beacon during their daily life. There are 4 different dialogue scenarios and a scenario is posted to the participants when he/ she approaches the interactive signage. Participants can respond to these scenarios through a touchscreen. Based on the result of the experiment, it is found that the interactive signage can induce behavior change effectively.

DOI： 10.23919/ICMU.2018.8653587

Language：English   Publishing type：Research paper (other academic)  

DOI： 10.1109/SEC.2018.00022

Language：Others   Publishing type：Research paper (other academic)  

DOI： 10.1145/3267305.3267667

Language：English   Publishing type：Research paper (other academic)  

© Copyright held by the owner/author(s). In this paper, we propose a battery-free place recognition system that utilizes solar-cells as a sensor for localization. Our system combines multiple solar cells having different characteristics against the light environment. As an initial work, we select five kinds of solar cells available in the market and investigate the characteristics of them. Then, we show a potential for estimating the place based on the variation of electricity amounts generated from solar cells. Finally, we show that our proposed system distinguish nine places with 88.0&#37; accuracy.

DOI： 10.1145/3267305.3267624

Language：English   Publishing type：Research paper (other academic)  

Recently, there is a growing demand to know con- gestion information on sightseeing spots in real-time to provide a satisfactory tour plan to tourists. Many studies on a congestion estimation have been conducted so far. However, most of them suffer from high deployment/operation costs and/or rely on contributions by users with smartphones/sensors. In this paper, we propose a novel system that estimates congestion of sightseeing spots in real-time without attaching any device to tourists by observing the distribution of per-RSSI intensity occurrences in each time window when beacon signals are periodically sent between BLE (Bluetooth Low Energy) transceivers installed in sightseeing spots. In other words, our system can estimate the congestion degree in sightseeing spots simply by using the property of RSSI intensity which dramatically changes depending on the number of people. Therefore, the proposed system is simple and low cost and can estimate the congestion easily without any special devices attached to tourists. In the demonstration, we show the congestion degree in three levels (low, medium, and high) changing in real-time depending on the number of audience in the demonstration site.

DOI： 10.1109/PERCOMW.2018.8480395

Language：English   Publishing type：Research paper (other academic)  

In this paper, we propose a battery-free place recognition system that utilizes solar-cells as a sensor for localization. Our system combines multiple solar cells having different characteristics against the light environment. As an initial work, we select five kinds of solar cells available in the market and investigate the characteristics of them. Then, we show a potential for estimating the place based on the variation of electricity amounts generated from solar cells. Finally, we show that our proposed system distinguish nine places with 88.0% accuracy.

DOI： 10.1145/3267305.3267624

Language：English   Publishing type：Research paper (other academic)  

In this paper, we, Ubi-NUTS Japan, introduce a multi-stage activity inference method that can recognize a user's mode of locomotion and transportation using mobile device sensors. We use the Sussex-Huawei Locomotion-Transportation (SHL) dataset to tackle the SHL recognition challenge, where the goal is to recognize 8 modes of locomotion and transportation (still, walk, run, bike, car, bus, train, and subway) activities from the inertial sensor data of a smartphone. We adopt a multi-stage approach where the 8 class classification problem is divided into multiple sub-problems considering the similarity of each activity. Multimodal sensor data collected from a mobile phone are inferred using a proposed pipeline that combines feature extraction and 4 different types of classifiers generated using the random forest algorithm. We evaluated our method using data from over 271 hours of daily activities of 1 participant and the 5-fold cross-validation. Evaluation results demonstrate that our method clearly recognizes the 8 types of activities with an average F1-score of 97%.

DOI： 10.1145/3267305.3267526

Language：English   Publishing type：Research paper (other academic)  

A body weight training (BWT) means the training which utilizes the self-weight instead of the weight machine. The feedback of form and proper training menu recommendation is important for maximizing the effect of BWT. The objective of this study is to realize a novel support system which allows beginners to perform effective BWT alone, under wearable computing environment. To make an effective feedback, it is necessary to recognize BWT type with high accuracy. However, since the accuracy is greatly affected by the position of wearable sensors, we need to know the sensor position which achieves the high accuracy in recognizing the BWT type. We investigated 10 types BWT recognition accuracy for each sensor position. We found that waist is the best position when only 1 sensor is used. When 2 sensors are used, we found that the best combination is of waist and wrist. We conducted an evaluation experiment to show the effectiveness of sensor position. As a result of leave-one-person-out cross-validation from 13 subjects to confirm validity, we calculated the F-measure of 93.5% when sensors are placed on both wrist and waist.

DOI： 10.1145/3267305.3267504

Language：English   Publishing type：Research paper (other academic)  

It has been intensively demonstrated that physical activity can enhance the mental and physical health of practitioners. In recent years, fitness activities became the most common way to engage in physical activities. In this paper, we propose a smart-glove based fitness activity tracking system that can detect athletes activities in any indoor fitness facility, with no need of attaching multiple sensors on the athlete's body. The system adopts force sensitive resistor (FSR) sensors to identify the type of exercise by analyzing the pressure distribution in the hand palm during fitness activities. To evaluate the performance of our proposed system, we ran a pilot study with 10 healthy participants over 10 common fitness activities. The experimental results showed an overall recognition accuracy rate of 87%. We believe the promising results would contribute to the works on personal assisted coaching systems and create enjoyable experiences when performing fitness activities.

DOI： 10.1109/PERCOMW.2018.8480211

Language：English   Publishing type：Research paper (other academic)  

To design more context-aware systems for smart environments, especially smart cities, the psychological user status such as emotion should be considered in addition to environmental information. In this study, we focus on the tourism domain as a typical use case, and propose a multimodal tourist emotion recognition method during the sightseeing. We employ behavioural cues (eye and head/body movements) and audio-visual features to recognise emotion. As a result of real-world experiments with tourists, we achieved up to 0.71 of average recall score in 3-class emotion recognition task with feature level fusion.

DOI： 10.1145/3267305.3267687

Language：English   Publishing type：Research paper (other academic)  

Recently, one-way car-sharing service that allows a user to drop the vehicle off at the destination has been started. However, it has not been as popular as expected mainly because it cannot assure the existence of an available vehicle at the start station or a vacant parking slot at the destination station. At this moment, the company staff are responsible for moving the vehicles to certain stations manually. Therefore, it is very costly. In order to reduce costs and make more efficient use of vehicles in one-way car-sharing service, we propose to involve a future potential user to move vehicles. We assume that there exist some requested one-way trips. Our method inserts a new trip among the existing trips for maximizing the number of successful trips. This trip is called 'a system request trip.' We assume that the system request trip can be realized by changing a future schedule or asking future potential users who want to get some rewards. In this paper, we analyzed how each user uses vehicles in one-way car-sharing service by using the past data of demonstration experiment conducted by a car-sharing company in Tokyo, Japan. Based on the analysis result, we examined how to request each user.

DOI： 10.1109/PERCOMW.2018.8480256

Language：English   Publishing type：Research paper (other academic)  

We introduce a Graphical User Interface (GUI) to create intelligent interactive documents for everyone. The intelligent interactive document refers to a document displaying contents dynamically according to a reader's behavior. To the best of our knowledge, creating such documents requires a certain amount of efforts and implementation skills. By utilizing our system, users including non-technological experts can create interactive documents without any programming. Our system supports many people to enhance the possibility of designing new human-document interactions.

DOI： 10.1145/3267305.3267667

Language：Japanese  

Language：Japanese  

Consideration of Walking Support Application Based on Music Entrainment Effect

Language：Japanese  

Request of the Manuscript for IPSJ Kansai-Branch Convention 2018

Language：English   Publishing type：Research paper (other academic)  

Emerging smart appliances provide us with the opportunity of understanding peoples' in-depth daily life behavior patterns, through analyzing their operation states. Smart fridges, for example, can help us estimate the timing and content of meals of household members. Several smart fridges already exist. However, they have the following problems: high cost, a risk of privacy leakage through camera, and inability of tracking incremental consumption of food-stuffs. In this paper, to address these problems, and to realize smooth in-fridge foodstuff management, we propose a smart fridge with a weight sensor and a voice interface. The fridge stores the name and weight of each food item in a database. When a user places an item, whose information is not already stored in the database, the system supports the user to easily register the information with the weight sensor and the voice interface. In addition, we propose three input methods for efficient association between foodstuff names and their weights: sequential, asynchronous parallel, and synchronous parallel methods. We developed a prototype of our proposed smart fridge system and conducted a preliminary experiment with three participants to register foodstuffs and compared the time for registration between the proposed method and a popular existing smartphone application for in-fridge foodstuff management. As a result, we confirmed that the proposed system can register the foodstuffs in hands-free with comparable time to the existing application.

DOI： 10.1145/3229710.3229727

Language：English   Publishing type：Research paper (other academic)  

DOI： 10.1145/3229710.3229727

Language：English   Publishing type：Research paper (other academic)  

Following the increase in demand for 'smart tourism,' various tourist information becomes available. Current tourist guidance systems can provide many possible routes around the city, which are usually based on an optimal distance and time or popularity of the place. To design more enjoyable tourist routes, we should be aware of tourist's perception of the urban environment. In this paper, we propose the methodology of estimating the tourist emotions and satisfaction level by analysing physiological features, such as head movement and eye gaze. Features derived from raw sensor data have the correlation up to 0.58 with emotion and satisfaction labels. This study also shows the differences in feature/label dependencies between different touristic areas and highlights the challenges of tourist satisfaction estimation.

DOI： 10.1109/SMARTCOMP.2018.00036

Language：English   Publishing type：Research paper (scientific journal)  

DOI： 10.1155/2018/2625195

Language：Japanese  

Language：English   Publishing type：Research paper (scientific journal)  

Heart Rate Prediction for Easy Waking Route Planning

Language：Japanese  

Language：English   Publishing type：Research paper (other academic)  

DOI： 10.1109/SMARTCOMP.2018.00036

Language：English   Publishing type：Research paper (other academic)  

DOI： 10.3233/978-1-61499-874-7-244

Language：English   Publishing type：Research paper (other academic)  

In an era where billions of IoT devices are deployed, edge/fog computing paradigms are attracting attention for their ability to reduce processing delays and mitigate waste of communication resources. However, since the computing system assumed by edge/fog paradigms have heterogeneity (in terms of the computing power of devices, network performance between devices, device density, etc.), provisioning computational resources according to computational demand becomes a challenging constrained optimization problem. In this paper, we propose in-situ resource provisioning method consisting of insitu resource area selection with adaptive scale out and in-situ task scheduling based on tabu search algorithm. We conducted a simulation study in a target regional area where 2,000 IoT devices and 10 IoT services are deployed to evaluate the effectiveness of the proposed algorithm. The simulation results show that our proposed algorithm can obtain higher user QoS compared to conventional resource provisioning algorithms.

DOI： 10.1109/IWQoS.2018.8624127

Language：English   Publishing type：Research paper (other academic)  

DOI： 10.3233/978-1-61499-874-7-196

Language：Japanese  

Language：Japanese  

Language：Japanese  

Language：Japanese  

Language：Japanese  

Language：Japanese  

Language：Japanese  

Language：Japanese  

Language：English   Publishing type：Research paper (other academic)  

DOI： 10.1007/978-981-13-9443-0_15

Language：Japanese  

Language：Japanese  

Language：Japanese  

Language：Japanese   Publishing type：Research paper (scientific journal)  

Language：English   Publishing type：Research paper (other academic)  

More than 60 percentage of fatal accidents while riding a bicycle is caused by elderly people over 65 years old. The main cause is the detection delay of approaching vehicle caused by the decrease of cognitive function due to aging. In this paper, we propose an approaching vehicle detection method using a smartphone aiming to support bicycle operation to prevent elderly people from fatal accidents while riding a bicycle vehicle. Among various sensors embedded in a smartphone, we focus on microphone as the most suitable sensor for detecting an approaching vehicle. We collected sound data in a real environment and created an approaching vehicle detection model by using machine learning. Finally, as a result of accuracy evaluation with 10-fold cross-validation, we confirmed that it can detect approaching vehicle with an average F-value of 97.4 [%].

DOI： 10.23919/ICMU.2017.8330069

Language：English   Publishing type：Research paper (other academic)  

Recognizing the activities of daily living (ADL) of residents in housing is indispensable for operating Daily Life Support Services such as Elderly Monitoring, Smart Home Automation, and Health Support. However, the existing methods have various problems: invasion of privacy, limited target activities, low recognition accuracy, initial installation cost, and long recognition time. As our prior work, we proposed a real-time ADL recognition method using indoor positioning sensor and power meters. We got a result that the method can recognize ten types of ADL with the average accuracy of 79%. However, the accuracy of some activities such as work/study and bathroom-related were not satisfactory. In this work, we aim to improve the accuracy of our prior method by newly adding several new time and are related features such as time slot when activity occurs, staying time in the same area, and previous position. As a result, we could achieve 82% of average recognition accuracy for 10 different activities.

DOI： 10.23919/ICMU.2017.8330104

Language：English  

Emerging Internet of Things (IoT) technologies will allow spectators in a sport game to produce various video streams from various angles. With existing technologies, however, it is difficult to process massive and various data streams for multi-channel contents in real-time. To solve this problem, we aim to construct a software agent (called "Curator") that compiles video contents automatically according to his/ her values. In this paper, we propose a system to automatically switch multiple video streams that general sports spectators have taken using Random Forests classifier. Meta data such as image feature data and game progress data is extracted for each video scene as the input of the classifier. For evaluation, we constructed a camera switching timing estimation model using the live TV broadcast of some baseball game data. A video of another baseball game was curated with the constructed model. As a result, our system predicted the camera switching timing with accuracy (F-measure) of 85.3% on weighted average for the base camera work and 99.7% for the fixed camera work.

DOI： 10.4018/978-1-5225-5475-2.ch022

Language：English   Publishing type：Research paper (other academic)  

Recently, there is a growing demand to know congestion information on sightseeing spots in real-time to provide a satisfactory tour plan to tourists. Many studies on a congestion estimation have been conducted so far. However, most of them suffer from high deployment/operation costs and/or rely on contributions by users with smartphones/sensors. In this paper, we propose a novel system that estimates congestion of sightseeing spots in real-time without attaching any device to tourists by observing the distribution of per-RSSI intensity occurrences in each time window when beacon signals are periodically sent between BLE (Bluetooth Low Energy) transceivers installed in sightseeing spots. In other words, our system can estimate the congestion degree in sightseeing spots simply by using the property of RSSI intensity which dramatically changes depending on the number of people. Therefore, the proposed system is simple and low cost and can estimate the congestion easily without any special devices attached to tourists. In the demonstration, we show the congestion degree in three levels (low, medium, and high) changing in real-time depending on the number of audience in the demonstration

DOI： 10.1109/PERCOMW.2018.8480395

Language：English   Publishing type：Research paper (other academic)  

DOI： 10.1109/PERCOMW.2018.8480211

Language：English   Publishing type：Research paper (other academic)  

DOI： 10.1109/PERCOMW.2018.8480256

Language：English   Publishing type：Research paper (other academic)  

DOI： 10.1109/PERCOMW.2018.8480348

Language：Japanese  

Language：Japanese  

Language：Japanese  

Language：Japanese  

Development and Evaluation of In-home Activity Recognition Utilizing Doppler Sensor

Language：Japanese  

Language：Japanese  

Language：English   Publishing type：Research paper (other academic)  

Development of Automatic Daily Report Generation System for Supporting Daily Works on Carers

Language：English   Publishing type：Research paper (other academic)  

Sitting on a chair with incorrect posture, or the same posture for a long time, causes the burden on the waist and shoulders of workers, leading to mental stress and poor work efficiency. It is therefore important to measure workers' sitting posture and advice on how to improve the posture or notify them to either adjust their posture or take a break from sitting. To achieve this, it is necessary to continuously measure how a worker is sitting without affecting the worker's ability to complete their tasks. We developed a sensing chair called 'Census' satisfying these requirements. This sensing chair has a function that is able to classify the sitting posture of workers with an accuracy of 80.1%.

DOI： 10.1109/ICCE.2018.8326275

Language：Japanese  

Language：English   Publishing type：Research paper (scientific journal)  

In this paper, we aiming to realize near-zero energy distributed IoT systems that do not need a battery, we design and implement a novel energy harvesting (EH) sensor node. The proposed EH sensor node consists of a solar panel, an energy charging/discharging circuit and a SenStick, an existing sensor node including 8 different types of sensors and BLE communication interface. We demonstrate that the proposed EH sensor node works in an indoor environment by harvesting power from indoor lights, periodically measuring sensor data (temperature, humidity, atmospheric pressure, UV, and illuminance) and by sending the data by BLE communication. The received information is visualized through a web application running on a PC.

DOI： 10.1007/978-3-319-90740-6_13

Language：Japanese  

Language：Japanese  

Tour Video Curation Method for Intuitive Sightseeing Tour Planning

Language：English   Publishing type：Research paper (other academic)  

Health-related quality of life (HRQOL) is a useful indicator that rates a person’s activities in various physical, mental and social domains. Continuously measuring HRQOL can help detect the early signs of declines in these activities and lead to steps to prevent such declines. However, it is difficult to continuously measure HRQOL by conventional methods, since its measurement requires each user to answer burdensome questionnaires. In this paper, we propose a simplified HRQOL measurement method for a continuous HRQOL measurement which can reduce the burden of questionnaires. In our method, sensor data from smart devices and the questionnaire scores of HRQOL are collected and used to construct a machine-learning model that estimates the score for each HRQOL questionnaire item. Our experiment result showed our method’s potential and found effective features for some questions.

DOI： 10.1007/978-3-319-98551-0_11

Language：English   Publishing type：Research paper (scientific journal)  

Smart mobility systems, which include Intelligent Transportation Systems (ITS) and smart energy ones, become more important. There is, however, lack of its platform studies. This paper proposes a sustainable information infrastructure project for smart mobility systems. The project pursues issues that establish an information infrastructure architecture and a seamless development method chain for it. The project has mainly two features: (1) applying life-cycle-oriented methods, which are a cycle from system development phase to operation phase. In addition, these methods are applied to real world applications, and (2) dealing with an uncertainty that occurs in system development upper phase. This paper describes an overview of the project.

DOI： 10.1007/978-3-319-70636-8_3

Language：English   Publishing type：Research paper (other academic)  

DOI： 10.1109/IIAI-AAI.2018.00025

Language：English   Publishing type：Research paper (scientific journal)  

DOI： 10.1109/ACCESS.2018.2793844

Language：English   Publishing type：Research paper (other academic)  

Sitting on a chair with incorrect posture, or the same posture for a long time, causes the burden on the waist and shoulders of workers, leading to mental stress and poor work efficiency. It is therefore important to measure workers' sitting posture and advice on how to improve the posture or notify them to either adjust their posture or take a break from sitting. To achieve this, it is necessary to continuously measure how a worker is sitting without affecting the worker's ability to complete their tasks. We developed a sensing chair called 'Census' satisfying these requirements. This sensing chair has a function that is able to classify the sitting posture of workers with an accuracy of 80.1%.

DOI： 10.1109/ICCE.2018.8326275

Language：Japanese  

Proposal of Rent Estimation Model for Restaurant Properties based on Tacit Knowledges
Previously determining the rent of restaurants properties was based on the tacit knowledge, that is, intuition and experience gained by experienced salespeople. However, this business custom has problems (e.g., no evidence to the determined rent). Therefore, the transference of the knowledge and experience from the experienced salespeople to fresh one, is not effective. To make rent estimation more correctly and efficiently, we build the rent estimation concept model for restaurants properties. In addition, we discuss about the specific challenges in building the rent estimation system, that are (1) acquisition and indexing of the tacit knowledge and (2) construction of a rent estimation model. In this paper, we build rent estimation models based on some machine-learning methods. As the result of comparing, the case of using Random Forest regression algorithm achieved the highest accuracy that coefficient of determination was 0.738.

Language：English   Publishing type：Research paper (scientific journal)  

TouchAnalyzer: A System for Analyzing User’s Touch Behavior on a Smartphone

Language：English  

Smart mobility systems, which include Intelligent Transportation Systems (ITS) and smart energy ones, become more important. There is, however, lack of its platform studies. This paper proposes a sustainable information infrastructure project for smart mobility systems. The project pursues issues that establish an information infrastructure architecture and a seamless development method chain for it. The project has mainly two features: (1) applying life-cycle-oriented methods, which are a cycle from system development phase to operation phase. In addition, these methods are applied to real world applications, and (2) dealing with an uncertainty that occurs in system development upper phase. This paper describes an overview of the project.

DOI： 10.1007/978-3-319-70636-8_3

Language：English   Publishing type：Research paper (scientific journal)  

Social support can improve adherence to physical activity and healthy eating behavior change, and recently, smartphone fitness applications which provide interventions to groups of users are increasingly popular. However, the types of social support provided in the applications, and the features/approaches used to offer the support, vary widely. It is unknown which approach is the best to meet users' requirements. We conducted a mixed methods study to identify user requirements for social support in group fitness applications. We then developed MyFitnessTeam based on these requirements and evaluated it in a six-week field study. We found that participants desired informational and emotional support, and for group fitness applications to provide daily challenges that they could complete with their support network. MyFitnessTeam significantly improved adherence to fitness behaviors, with men having significantly higher adherence levels than women and participants who were inactive at baseline achieving similar adherence levels to the most baseline-Active participants. Groups of friends were more likely to exchange informational support but emotional support was low for both friends and strangers, due to lack of knowledge on when and how to offer such support. We propose five design recommendations to promote supportive interactions in group fitness applications.

DOI： 10.1109/ACCESS.2018.2793844

Language：English   Publishing type：Research paper (other academic)  

In this paper, we aiming to realize near-zero energy distributed IoT systems that do not need a battery, we design and implement a novel energy harvesting (EH) sensor node. The proposed EH sensor node consists of a solar panel, an energy charging/discharging circuit and a SenStick, an existing sensor node including 8 different types of sensors and BLE communication interface. We demonstrate that the proposed EH sensor node works in an indoor environment by harvesting power from indoor lights, periodically measuring sensor data (temperature, humidity, atmospheric pressure, UV, and illuminance) and by sending the data by BLE communication. The received information is visualized through a web application running on a PC.

DOI： 10.1007/978-3-319-90740-6_13

Language：English   Publishing type：Research paper (scientific journal)  

As the world's population of senior citizens continues to grow, the burden on the professionals who care for them (carers) is also increasing. In nursing homes, carers often write daily reports to improve the resident's quality of life. However, since each carer needs to simultaneously care for multiple residents, they have difficulty thoroughly recording the activities of residents. In this paper, we address this problem by proposing an automatic daily report generation system that monitors the activities of nursing home residents. The proposed system estimates the multiple locations (areas) at which residents are situated with a BLE beacon, using a variety of methods to analyze the RSSI values of BLE signals, and recognizes the activity of each resident from the estimated area information. The information of the estimated activity of residents is stored in a server with timestamps, and the server automatically generates daily reports based on them. To show the effectiveness of the proposed system, we conducted an experiment for five days with four participants in cooperation with an actual nursing home. We determined the proposed system's effectiveness with the following four evaluations: (1) comparison of performance of different machine-learning algorithms, (2) comparison of smoothing methods, (3) comparison of time windows, and (4) evaluation of generated daily reports. Our evaluations show the most effective combination pattern among 156 patterns to accurately generate daily reports. We conclude that the proposed system has high effectiveness, high usability, and high flexibility.

DOI： 10.1155/2018/2625195

Language：English   Publishing type：Research paper (scientific journal)  

DOI： 10.1007/978-3-319-98551-0_11

Language：Japanese  

Language：English   Publishing type：Research paper (other academic)  

In this paper, we investigate an effect on group detection based on human proximity depending on RSSI values, measured by BLE signals. To realize the investigation, we present a novel method for detecting groups. In the method, we developed a smartphone application for measurement of RSSI values of BLE, and determine a threshold of RSSI values which selected experimentally by making the propagation characteristics of BLE signal. To show the effectiveness of proposed method, we evaluated the method through a series of experiments and investigated the effect on group detection by the difference of human proximity. As the results, we found that the group detection was affected by the difference of human proximity.

DOI： 10.1109/ICPPW.2017.14

Language：Japanese  

A method for highlights extraction toward tour video curation

Language：Japanese  

Language：Japanese  

Type Recognition by Wearable Devices for Effective Personal Workout

Language：Japanese  

Language：Japanese  

Language：Japanese  

Language：Japanese  

A Design of the System Triggered Behavior Change Based on Focusing Sensing by Eyewear

Language：Japanese  

Analysis method for formalization of thought process at web search

Language：English   Publishing type：Research paper (scientific journal)  

Smartphone-based Food Weight and Calorie Estimation Method for Effective Food Journaling
World health statistics about overweight and obesity show that overweight, obesity and diet-related diseases still remain major health risks. According to the World Health Organization, most of the world's population live in countries where overweight and obesity kill more people than underweight. Recently, many studies have been driven by the motivation of elaborating the &ldquo;ideal&rdquo; solutions to prevent and/or monitor overweight, obesity and diet-related diseases, and to encourage healthy diet and lifestyle. In this work, we present our idea to automatically measure food weight and calories, from food photo using ordinary chopsticks as a measurement reference. The analysis of the obtained results show that the use of near-by eating utensils combined with computer vision techniques is a great and exploitable approach to ubiquitously help in diet assessment and obesity treatment.

DOI： 10.9746/jcmsi.10.360

Language：English   Publishing type：Research paper (scientific journal)  

Implementation and Evaluation of Analog-PIR-sensor-based Activity Recognition
These days, smart home applications such as concierge service for residents, home appliance control, and so on are attracting attention. To realize these applications, we need a system which recognizes various human activities accurately with a low cost device. There are many studies on the activity recognition in a smart home. We also have proposed an activity recognition technique in a smart home by utilizing digital-output-PIR (passive infrared) sensors, door sensors, and power meters. However, the study has an unsolved issue: we cannot distinguish similar activities happening at the same place, for example, &ldquo;eating&rdquo; and &ldquo;reading&rdquo; while sitting on a sofa. In order to cope with this challenge, we introduce ALPAS: analog-output-PIR-sensor-based activity recognition technique which recognizes the different activities in the same place. Our technique recognizes user's activity by utilizing machine learning with frequency components of the sensor's output as features. However, because the number of features used in ALPAS is 1000 for each analog PIR-sensor, a large capacity memory is required. To reduce the number of features, we select a part of the sensing data. We call the starting point of the selected data as starting frequency (SF) and ending part as ending frequency (EF). We searched SF and EF using a grid search, and evaluated the recognition accuracy. We evaluated the proposed technique in a smarthome testbed. In the evaluation, five participants performed four different activities while sitting on a sofa. As a result, we achieved F-Measure: 63.9&#37; when the EF is 1.4Hz, and F-Measure: 50&#37; or lower when the SF is 9.9Hz or higher.

DOI： 10.9746/jcmsi.10.385

Language：English   Publishing type：Research paper (other academic)  

This paper proposes a tool for tracking a thought process in web searching because we considered that user's browsing history potentially contains the transition information of a thought process of how to collect information efficiently. In fact, we have few opportunities to look back own browsing way. Even if we check the enormous browsing logs, it is hard to convert them into sharable knowledge. Therefore, our tool measure the dwell time, the amount of scrolling and page contents and transition between pages for evaluating the importance of each page. In addition, our tool provides an editing function to summarize a thought process based on the importance of each page. Copyright held by the owner/author(s).

DOI： 10.1145/3123024.3129267

Language：English   Publishing type：Research paper (other academic)  

A System for Collecting and Curating Sightseeing Information toward Satisfactory Tour Plan Creation

Language：English   Publishing type：Research paper (other academic)  

Recommending Optimal Route for Walking Support based Heart Rate Prediction

Language：Japanese  

Language：Japanese  

Language：Japanese  

Language：Japanese  

Language：Japanese  

Language：English   Publishing type：Research paper (other academic)  

These days, smart home applications such as a concierge service for residents, home appliance control and so on are attracting attention. In order to realize these applications, we strongly believe that we need a system which recognizes the various human activities accurately with a low cost device. There are many studies which work on the activity recognition in the smarthome. Moreover, we also have proposed the activity recognition technique in the smarthome by utilizing the digital-output-PIR sensor, door sensor, watt meter. However, the study has the challenge: we cannot distinguish between the similar tiny activities at the same place: 'eating' and 'reading' with sitting on a sofa. In order to cope with this challenge, we introduce ALPAS: analog-output-PIR-sensor-based activity recognition technique which recognizes the detailed activities of the user. Our technique recognizes the activity of the user by utilizing the machine learning. We evaluated the proposed technique in a smarthome which belongs to the authors' university. In the evaluation, three subjects performed four different activities with sitting on a sofa. As a result, we achieved F-Measure: 57.0%.

DOI： 10.1109/AINA.2017.33

Language：Japanese  

Implementation and Evaluation of a Participatory Location Sharing System for Cycling Road Race with Reducing Burden for the Athletes
Athlete's location sharing service has already started in various cycling events. However, in such system, the athletes have to bring a smartphone for uploading their position in racing. In this paper, we propose a participatory location sharing system that utilizes supporters' smartphone for tracking the position of athletes. In our system, the athletes only have a small BLE beacon, and the supporters along the road have a smartphone in which our application is installed. The application uploads and shares its location when it receives a beacon signal. Since there is no existing work that try to track a fast-passing beacon (the speed of bicycle is 30km/h-80km/h), we developed the dedicated beacon and dedicated Android application for experiment. Through various experiments, we clarified the effect of passing speed, the maximum distance for receiving a signal, location accuracy compared with GPS, the effect of landform and shielding. Finally, we conducted an experimental evaluation in the actual cycling events for confirming the availability and effectiveness.

Language：Japanese  

Language：Japanese  

Language：Japanese  

Language：English   Publishing type：Research paper (other academic)  

Automatic recognition of activities of daily living (ADL) can be applied to realize services to support user life such as elderly monitoring, energy-saving home appliance control, and health support. In particular, 'real-time' ADL recognition is essential to realize such a service that the system needs to know the user's current activity. There are many studies on ADL recognition. However, none of these studies address all of the following problems: (1) privacy intrusion due to the utilization of high privacy-invasive devices such as cameras and microphones; (2) limited number of recognizable activities; (3) low recognition accuracy; (4) high deployment and maintenance costs due to many sensors used; and (5) long recognition time. In our prior work, we proposed a system which solves the problems (1)-(4) to some extent by using user's position data and power consumption data of home electric appliances. In this paper, aiming to solve all the above problems including (5), we propose a new system by extending our prior work. To realize 'real-time' ADL recognition while keeping good recognition accuracy, we developed new power meters with higher sensing frequency and introduced new techniques such as adding new features, selecting the best subset of the features, and selecting the best training dataset used for machine learning. We collected the sensor data in our smart home facility for 11 days, and applied the proposed method to these sensor data. As a result, the proposed method achieved accuracy of 79.393% in recognizing 10 types of ADLs.

DOI： 10.1109/PERCOMW.2017.7917620

Language：English   Publishing type：Research paper (other academic)  

Recently, as elderly people population grows, the burdens on caretakers are getting larger. In daycare centers, caretakers make a daycare report aiming to improve the senior citizen's Quality of Life. However, in the present situation, it is difficult for caretakers to record the senior citizen's activity in detail, since each caretaker needs to take care of several senior citizens at the same time. To reduce the burden of caretakers, many elderly monitoring systems have been proposed so far, but most of them are not effective in the sense that they force the senior citizen to use dedicated devices such as smart phone and/or particular applications that are obtrusive and cumbersome for care receivers. In this paper, we propose a semi-automatic care-taking report generation system which can monitor movements/activity of senior citizens in daycare centers. Our proposed system estimates multiple locations (areas) where senior citizens are located with the BLE beacon, by utilizing RSSI of the Bluetooth radio wave. Also, the accelerometer implemented in the tag estimates the activity of the elderly. The information of the estimated area and activity is stored in a server with time stamp. The server generates the daycare report based on it. In order to evaluate the proposed system, we have deployed our system in a daycare center: Ikoi-no-ie 26. Evaluation result in Ikoi-no-ie 26 showed that our system estimated the subject's present area with F-measure: 80.6% and activity with F-measure: 73.8% and generated the daycare report.

DOI： 10.1109/AINA.2017.34

Language：English   Publishing type：Research paper (other academic)  

Companies have recently been introducing an event system called shuffle lunch, which aims to effectively utilize lunch breaks. The system has been garnering attention from many enterprises because it livens relationships and strengthens cooperation between departments and individual workers. In existing shuffle lunch systems, lunch groups are randomly generated. Random groups however, can cause problems. For example, some people might feel awkward about eating with new people, groups might not agree about when and where to eat, or some people may be absent. To form better lunch groups, the relationships between potential group members have to be known. In this study, we propose a method for dynamically detecting groups by using smartphones to measure the daily physical proximities of people. We also developed an Android application to realize our proposed method. We evaluated our system through a series of experiments and found that our proposed method can accurately detect groups, based on the proximities measured by the Android application.

DOI： 10.1109/WAINA.2017.43

Language：English   Publishing type：Research paper (other academic)  

This paper presents a study about estimating the emotions conveyed in clips of background music (BGM) to be used in an automatic slideshow creation system. The system we aimed to develop, automatically tags each given pieces of background music with the main emotion it conveys, in order to recommend the most suitable music clip to the slideshow creators, based on the main emotions of embedded photos. As a first step of our research, we developed a machine learning model to estimate the emotions conveyed in a music clip and achieved 88% classification accuracy with cross-validation technique. The second part of our work involved developing a web application using Microsoft Emotion API to determine the emotions in photos, so the system can find the best candidate music for each photo in the slideshow. 16 users rated the recommended background music for a set of photos using a 5-point likert scale and we achieved an average rate of 4.1, 3.6 and 3.0 for the photo sets 1, 2, and 3 respectively of our evaluation task.

DOI： 10.1109/PERCOMW.2017.7917587

Language：English   Publishing type：Research paper (other academic)  

Recently, IoT (Internet of Things) technologies have been attracting increasing attention. Among many applications of IoT, homes can be the most promising target. One of the purposes to deploy IoT in homes is automatic recognition of activities of daily living (ADLs). It is expected that ADL recognition in homes enables many new services such as elderly people monitoring and low energy appliance control. In existing studies on ADL recognition, however, it is hard to build a system to acquire data for ADL recognition in terms of installation cost. In this paper, we propose a method that reduces costs of the ADL recognition system by using ECHONET Lite-ready appliances which are expected to be widely spread in the future. ECHONET Lite is a communication protocol for control and sensor networks in smart-homes and standardized as ISO/IEC-4-3. The proposed method utilizes information (e.g., on/off state) from appliances and motion sensors attached to them as features and recognizes ADLs through machine learning. To evaluate the proposed method, we collected data in our smart-home testbed while several participants are living there. As a result, the proposed method achieved about 68% classification accuracy for 9 different activities.

DOI： 10.1109/PERCOMW.2017.7917603

Language：English   Publishing type：Research paper (other academic)  

This paper shows the concept and design of an ongoing project about "behavior change", that is one of the keywords for realizing a sustainable society based on information technology such as IoT and AI. "Stand" function that Apple Watch has is a touchstone of the arrival of a new age. The watch commands a human to stand up or meditate. It is a start of intervention to our behavior from AI. However, it is not so bad because we know that this suggestion must be good for our health. Based on this experience, we started an empirical research on behavior change that consists of activity recognition, just-in-time intervention, and gamification.

DOI： 10.1145/3055031.3055068

Language：English   Publishing type：Research paper (scientific journal)  

Generating pedestrian maps of disaster areas is an important part of response operations. Maps aid responders in decision-Making and show routes that read evacuees to refuges. However, disasters can damage communication infrastructures, rendering Cloud-based mapping services inaccessible. Responders resort to paper maps, which are difficult to share and cannot recommend routes. In this study, we present a digital pedestrian map generation system for disasters. To realize the system, we addressed these challenges: (1) how to collect the required data and generate the map without Cloud-based computing resources, (2) how to share messages within the system without continuous, end-to-end networks, and (3) how to balance the load of map inference tasks. For (1), GPS traces are Collected by responders exploring the area. Then, collected data are sent to Computing Nodes: commodity workstations that are deployed in the disaster area, for processing. For (2), the system establishes a Delay-Tolerant Network that uses Epidemic Routing to communicate across shorter-ranges and uses response vehicles as data ferries to communicate across longer-ranges. For (3), we propose a load balancing heuristic, which uses ferry route timetables and statistical information about the load of Computing Nodes to determine how to offload map inference tasks. We evaluate our system through experiments and simulations and show that it decreases the time needed to generate and deliver pieces of the map by approximately two hours in an extreme case with large quantities of data have to be processed. (C) 2016 Elsevier B.V. All rights reserved.

DOI： 10.1016/j.comcom.2016.12.003

Language：English   Publishing type：Research paper (scientific journal)  

Generating pedestrian maps of disaster areas is an important part of response operations. Maps aid responders in decision-making and show routes that lead evacuees to refuges. However, disasters can damage communication infrastructures, rendering Cloud-based mapping services inaccessible. Responders resort to paper maps, which are difficult to share and cannot recommend routes. In this study, we present a digital pedestrian map generation system for disasters. To realize the system, we addressed these challenges: (1) how to collect the required data and generate the map without Cloud-based computing resources, (2) how to share messages within the system without continuous, end-to-end networks, and (3) how to balance the load of map inference tasks. For (1), GPS traces are collected by responders exploring the area. Then, collected data are sent to Computing Nodes: commodity workstations that are deployed in the disaster area, for processing. For (2), the system establishes a Delay-Tolerant Network that uses Epidemic Routing to communicate across shorter-ranges and uses response vehicles as data ferries to communicate across longer-ranges. For (3), we propose a load balancing heuristic, which uses ferry route timetables and statistical information about the load of Computing Nodes to determine how to offload map inference tasks. We evaluate our system through experiments and simulations and show that it decreases the time needed to generate and deliver pieces of the map by approximately two hours in an extreme case with large quantities of data have to be processed.

DOI： 10.1016/j.comcom.2016.12.003

Language：Japanese  

Language：Japanese  

Effect on tracking coverage of high power BLE beacon in participatory position sharing system for bicycling sports

Language：Japanese  

近年，デスクワークの増加により，椅子に座って作業するオフィスワーカーが増加している．オフィスワーカーの生産性や作業効率の向上のためには，ワークプレイスにおいて快適に作業に取り組めるようにすることが重要である．本研究では，椅子に座って作業することが多いオフィスワーカーに着目し，着座姿勢によるワークプレイスの高機能化を提案する．高機能化のためにモーションセンサを取り付けたオフィスチェアによって着座姿勢を推定し，その結果から長時間同じ姿勢で座っているときは休憩の通知，離床時にはPCの画面を消灯，椅子に深くもたれたときはコーヒーを自動で淹れるなどを行う．

Language：Japanese  

Language：Japanese  

Language：Japanese  

Sitting Posture Estimation Method Using a Chair with Motion Sensors

Language：Japanese  

Language：Japanese  

Language：Japanese  

Language：Japanese  

Language：Japanese  

近年、Google Maps等の地図サービスの普及に伴い、ユーザが興味のある場所(POI：Point of Interest)をまとめたマイマップの共有が行われている。2014年には、マイマップをフリーキーワードで検索可能なサービスが開始している。しかし、大半のマイマップに登録されているPOIにはカテゴリ情報がなく、どのようなカテゴリのPOIをまとめマイマップか分からない現状にある。そこで、本稿ではPOIの名称からカテゴリを推定する方法を提案する。提案手法では、Foursquareから得られるPOIの名称とカテゴリの関係を機械学習させ、POIのカテゴリを推定する。評価の結果、25種類のカテゴリを55.6%の精度で推定可能である。また、上位5種の平均精度は82.0%、上位10種の平均精度は72.2%である。

Language：Japanese  

Language：Japanese  

Language：Japanese  

Language：Japanese  

Language：Japanese  

Language：Japanese  

Language：Japanese  

A Living Activity Recognition System Based on Power Consumption of Appliances and Energy Harvesting PIR and Door Sensors
Progress of ubiquitous computing technology has strong anticipation to realize efficient energy-saving appliance control and elderly monitoring. In order to put those applications into practice, high-accuracy and low-cost in-home living activity recognition is essential. Many researches have tackled living activity recognition so far, but the followings problems remain: (1) privacy exposure due to utilization of cameras and microphones; (2) high deployment and maintenance costs due to many sensors used; (3) burden to force the user to carry the device and (4) wire installation to supply power and communication between sensor node and server; (5) few recognizable activities; (6) low recognition accuracy. In this paper, we propose an in-home living activity recognition method to solve all the problems. To solove the problems (1)-(4), our method only utilizes energy harvesting PIR and door sensors and power meter. The energy harvesting sensor has a solar cell to drive the sensor and wireless communication modules. To solve the problems (5) and (6), we have tackled the following challenges: (i) determining appropriate features for training samples; and (ii) determining the best machine learning algorithm to achieve high recognition accuracy; (iii) complement for the dead zone of PIR sensor semipermanently. We have conducted evaluations with the sensor by five subjects living in a home for 2-3 days each. As a result, the proposed method has achieved F-measure: 68.6% on average.

Language：Japanese  

Language：English   Publishing type：Research paper (scientific journal)  

In this paper, we propose a recommendation model of application customer to reduce the time and cost of the salesperson by recommending the application customer to salesperson. The model is built based on customer information, premise information, and two new features which are extracted from salesperson’s feedback. Estimation precision is evaluated by three algorithms: SVM, Decision Tree and Random Forest. We applied three algorithm against five data sets. As a result, the highest estimation precision was 49.2% using RF against the data which selected by some important features (data 5). In the case of using the customer information and property information (data 3), estimation precision was 38.7%. Moreover in the case of adding new two features against that data (data 4), estimate precision was 44.2%. From comparison between data 3 and data 4, we clarified that the new two features increased the estimation precision. Also, from comparison between data 4 and data 5, we clarified that selecting some important features increased the estimation precision. In addition, according to the feedback from the ABC TENPO Inc which we are conducting joint research, our model increased the precision compared with the veteran salesperson.

DOI： 10.1527/tjsai.WII-O

Language：Japanese  

Research on Low Power Data Processing Technology in the network

Language：English   Publishing type：Research paper (scientific journal)  

We propose a comprehensive sensing platform called SenStick, which is composed of hardware (ultra tiny all-in-one sensor board), software (iOS, Android, and PC), and 3D case data. The platform aims to allow all the researchers to start IoT research, such as activity recognition and context estimation, easily and efficiently. The most important contribution is the hardware that we have designed. Various sensors often used for research are embedded in an ultra tiny board with the size of 50 mm (W) × 10 mm (H) × 5 mm (D) and weight around 3 g including a battery. Concretely, the following sensors are embedded on this board: acceleration, gyro, magnetic, light, UV, temperature, humidity, and pressure. In addition, this board has BLE (Bluetooth low energy) connectivity and capability of a rechargeable battery. By using 110 mAh battery, it can run more than 15 hours. The most different point from other similar boards is that our board has a large flash memory for logging all the data without a smartphone. By using SenStick, all the users can collect various data easily and focus on IoT data analytics. In this paper, we introduce SenStick platform and some case studies. Through the user study, we confirmed the usefulness of our proposed platform.

DOI： 10.1155/2017/6308302

Language：English   Publishing type：Research paper (other academic)  

Recently, as elderly people population grows, the burdens on caretakers are getting larger. In daycare center, caretakers make a daycare report aiming to improve the senior citizen's Quality of Life. However, in the aged society, it is difficult for caretakers to record the senior citizen's activity in detail, since each caretaker needs to take care of several senior citizens at the same time. In this paper, we propose a semi-automatic daycare report generation system. Our proposed system estimates the present area and activity of the senior citizen by utilizing the accelerometer-implemented-BLE-beacon tags and generate the daycare report. Based on the generated report, the caretaker works effectively. Evaluation result in a daycare center showed that our system estimated the subject's present area with F-measure: 80.6&#37; and activity with F-measure: 73.8&#37; and automatically generated the daycare report.

DOI： 10.1145/3012709.3016072

Language：English   Publishing type：Research paper (other academic)  

Recently, as elderly people population grows, the burdens on caretakers are getting larger. In daycare center, caretakers make a daycare report aiming to improve the senior citizen's Quality of Life. However, in the aged society, it is difficult for caretakers to record the senior citizen's activity in detail, since each caretaker needs to take care of several senior citizens at the same time. In this paper, we propose a semi-automatic daycare report generation system. Our proposed system estimates the present area and activity of the senior citizen by utilizing the accelerometer-implemented-BLE-beacon tags and generate the daycare report. Based on the generated report, the caretaker works effectively. Evaluation result in a daycare center showed that our system estimated the subject's present area with F-measure: 80.6% and activity with F-measure: 73.8% and automatically generated the daycare report.

DOI： 10.1145/3012709.3016072

Language：Japanese  

Language：Japanese  

Language：Japanese  

Language：English   Publishing type：Research paper (other academic)  

Thanks to rapid advance and penetration of IoT devices, it is becoming possible to sense almost every information of real-world. This urges us to utilize data streams continuously generated from IoT devices in real-Time. In this paper, aiming to locally process data streams by using computational resources of IoT devices, we propose middleware for IoT devices where the devices process data streams in real-Time and in a distributed manner. The proposed middleware provides four functions: (1) distribution of tasks issued by application software into sub-Tasks and distributed execution of the sub-Tasks over multiple IoT devices, (2) distribution of data streams over IoT devices, (3) real-Time analysis of the data streams, and (4) seamless integration of sensors and actuators. We have implemented a prototype of the proposed middleware for Raspberry Pi and show its basic performance.

DOI： 10.1109/ICDCSW.2016.37

Language：Japanese  

Language：Japanese  

Language：Japanese  

Language：English   Publishing type：Research paper (other academic)  

In this paper, we propose a health support system «WaistonBelt 2,» which consists of a belt-type wearable device and a smartphone application. Our system aims to support users to obesity (the metabolic syndrome) that may lead to various adult diseases such as arteriosclerosis, heart disease or stroke. Our proposed device can automatically measure a circumference on the belt position when a user wears it by a rotary encoder, and monitor a tension of belt for recognizing posture and overeating. The measured data is transmitted to our application via Bluetooth. Stored values are visualized by graphs, and advices are given through the avatar based on the analyzed data. Through evaluation experiments, we confirmed that our proposed device achieves highly accurate measurement of the circumference on the belt position (average absolute error is 0.73 cm). We also confirmed that it is possible to realize a recognition system of posture and eating action using belt pressure sensing system.

DOI： 10.1109/ICMU.2016.7742089

Language：English   Publishing type：Research paper (other academic)  

In this paper, we propose a novel rapid prototyping platform called SenStick, which is composed of both hardware and software. The main purpose of our platform is to sensorize our personal belongings easily and smart for recognizing our living activities. The most impressive point is its size. The size is 75mm(W) x 10mm(H) x 5mm(D) and its weight is around 3 (g) including a battery. On this tiny board, 8 sensors (acceleration, gyro, magnetic, light, UV, temperature, humidity, and pressure), flash memory, BLE, and battery are embedded in high density. The battery life in stand-alone mode is more than 12 hours. Second interesting point is the support software for iOS and Android OS. It can monitor the sensing data as well as can record the ground truth video simultaneously and synchronously. Furthermore, 3D CAD data of various case designs will be open for users in SenStick community site. As a result, SenStick enables everyone to sense every activities easily and smart.

DOI： 10.1109/ICMU.2016.7742087

Language：English   Publishing type：Research paper (other academic)  

In this paper, we propose an ad-hoc communication system for a smartphone, which utilizes Bluetooth Low Energy (BLE). The proposed system is assumed to be used in disaster situation where the Internet connection is not available. It means that a smartphone cannot obtain any information as well as any new application because all the applications are distributed in the cloud such as App Store. However, our proposed system is equipped with the mechanism, called RecurShare, to distribute the applications on-site. In addition, our proposed system has the function to solve «Central-Peripheral problem» which stems from BLE hardware. Generally, a smartphone works as «Central» of BLE connection, and the connection between Centrals cannot be established. Therefore, it is not easy to realize a BLE-based ad-hoc communication system between smartphones except iPhone. Our proposed function utilizes a few peripheral-ready terminal as a router of the network for solving this problem. To evaluate the feasibility of our proposed system, we show that the proposed system can replicate the application on-site and also enable chat among Android smartphones even if some of them are Central-only terminal.

DOI： 10.1109/ICMU.2016.7742096

Language：English   Publishing type：Research paper (other academic)  

During the recovery period after disasters, Family Tracing and Reunification (FTR) is the process by which separated family members are reunited. Traditional FTR methods rely on paper-based registries and notice boards, which cannot automatically match missing person queries with existing records and cannot be efficiently disseminated. Furthermore, lost children or people with disabilities may not be capable of supplying the text-based personal information required by registry forms. Finally, current digital FTR systems require the Internet for data delivery and storage, which may be unavailable during a disaster scenario. To overcome these limitations, we propose the Milk Carton FTR system. Milk Carton uses the Eigenfaces face recognition algorithm to automatically match missing person queries with existing records, without requiring text-based personal information. The system uses Computing Nodes, commodity devices deployed in evacuation centers, to handle record and query creation, data storage, and matching. To handle data delivery without the Internet, Milk Carton leverages response team vehicles as data ferries. The ferries store, carry, and forward records and queries across the system. In this study, we present the design of the Milk Carton system and initial performance evaluations.

DOI： 10.1145/3004010.3006380

Language：English   Publishing type：Research paper (other academic)  

Progress of IoT and ubiquitous computing technologies has strong anticipation to realize smart services in households such as efficient energy-saving appliance control and elderly monitoring. In order to put those applications into practice, high-accuracy and low-cost in-home living activity recognition is essential. Many researches have tackled living activity recognition so far, but the following problems remain: (i)privacy exposure due to utilization of cameras and microphones; (ii) high deployment and maintenance costs due to many sensors used; (iii) burden to force the user to carry the device and (iv) wire installation to supply power and communication between sensor node and server; (v) few recognizable activities; (vi) low recognition accuracy. In this paper, we propose an in-home living activity recognition method to solve all the problems. To solve the problems (i){(iv), our method utilizes only energy harvesting PIR and door sensors with a home server for data collection and processing. The energy harvesting sensor has a solar cell to drive the sensor and wireless communication modules. To solve the problems (v) and (vi), we have tackled the following challenges: (a) determining appropriate features for training samples; and (b) determining the best machine learning algorithm to achieve high recognition accuracy; (c) complementing the dead zone of PIR sensor semipermanently. We have conducted experiments with the sensor by five subjects living in a home for 2{3 days each. As a result, the proposed method has achieved F-measure: 62.8% on average.

DOI： 10.1145/3004010.3006378

Language：English   Publishing type：Research paper (other academic)  

Recently, there are many studies on automatic recognition of activities of daily living (ADL) to provide various ser-vices such as elderly monitoring, intelligent concierge, and health support. In particular, real-time ADL recognition is essential to realize an intelligent concierge service since the service needs to know user's current or next activity for supporting it. We have been studying real-time ADL recognition using only user's position data and appliances' power consumption data which are considered to include less privacy information than audio and visual data. In the study, we found that some activities such as reading and operating smartphone that happen in similar conditions cannot be classified with only position and power data. In this paper, we propose a new method that adds the acceleration data from wearable devices for classifying activities happening in similar conditions with higher accuracy. In the proposed method, we use the acceleration data from a smart watch and a smartphone worn by user's arm and waist, respectively, in addition to user's position data and appliances' power consumption data, and construct a machine learning model for recognizing 15 types of target activities. We evaluated the recognition accuracy of 3 methods: our previous method (using only position data and power consumption data); the proposed method using the mean value and the standard deviation of the acceleration norm; and the pro-posed method using the ratio of the activity topics. We collected the sensor data in our smart home facility for 12 days, and applied the proposed method to these sensor data. As a result, the proposed method could recognize the activities with 57 % which is 12 % improvement from our previous method without acceleration data.

DOI： 10.1145/3004010.3004036

Language：English   Publishing type：Research paper (other academic)  

In this paper, we propose an indoor localization method using a distance-illuminance model of lighting devices and trilateration. We propose a method that estimates distance from three controllable lighting devices based on the illuminance at the target point, by alternately turning on each of the devices. Then, the proposed method estimates the position of the target point based on trilateration. The proposed method are two merits. First, it can be realized at low cost because only three lighting devices and an illuminance sensor are required. Second, it is robust against influences by external lighting devices (or sunlight) since the proposed method measures the difference of the illuminance before and after turning on a lighting device. We conducted experiments in a room in an ordinary home environment and confirmed that the proposed method could estimate the position of the illuminance sensor within 0.5m error on average.

DOI： 10.1109/IPIN.2016.7743623

Language：English   Publishing type：Research paper (other academic)  

Consistent food tracking is important for weight loss. Many people struggle with forming and maintaining the habit because it is time consuming and generally un-enjoyable. We describe the design and initial field evaluation of HappyInu, a virtual pet-care game that we developed to promote consistent food tracking. HappyInu uses daily notifications to remind users to track, provides photo-based food tracking to make the process simpler and gives virtual and cash rewards for motivation. In our pilot study, users were excited about the game-based approach and they wanted more in-game tasks and rewards to be offered. Intrinsic rewards were more motivating than cash, particularly pleasure (fun) and the need to contribute to something meaningful. However, users played HappyInu on a casual (non-daily) basis, just as they did with other smartphone games. In this paper, we discuss users' perceptions on the role of games for behavior change and the implications on game design.

DOI： 10.1109/ICMU.2016.7742088

Language：English   Publishing type：Research paper (other academic)  

These days, smart home applications such as a concierge service for residents, home appliance control and so on are attracting attention. In order to realize these applications, we strongly believe that we need an indoor positioning system which fulfills the following requirements: Req 1: high accuracy; Req 2: low installation cost; Req 3: small burden on the user; Req 4: low privacy invasion. There are several studies which work on the indoor positioning system. However, these previous works do not accomplish the requirements. In this paper, we present a piezo sensor-based indoor positioning system which estimates the position of the user by utilizing a piezo component attached on the floor. To realize the proposed positioning system, we have tackled two challenges. First challenge is the development of an indoor positioning technique. We cannot utilize TDoA technique that is used to estimate the distance from the target, since the calculation of vibration velocity is difficult. To cope with this challenge, we have developed a new technique which estimates the position of the user from floor vibrations caused by their actions. Second challenge is the selection of the feature vector to estimate the vibration type accurately. We have selected MFCC, FFT, and Envelope shape features from preliminary experiments. We have implemented the proposed system in our smart home testbed. We have evaluated the performance of the vibration type estimation technique. As a result, we have confirmed that our technique estimates the type with F-measure: 93.9%.

DOI： 10.1109/IPIN.2016.7743667

Language：English   Publishing type：Research paper (other academic)  

Sensing thermal stress at office workplaces

Language：Japanese  

Language：Japanese  

Language：Japanese  

Language：English   Publishing type：Research paper (other academic)  

To increase the productivity of people's work, one of the most important challenges is to proactively keep and improve their mental health in the office while continuously obtaining their physiological information. In this paper, aiming to allow an office-worker to recognize his/her mental health and do self-care easily for prevention and reduction of his/her stress, as a pilot study, we propose Unconscious Stress Monitoring System for Office Worker, called COSMS. COSMS is a system acquiring physiological information continuously and unconsciously through ordinary equipment of a laptop such as in-camera and mouse. By using a laptop which embeds a depth camera, COSMS monitors a user's vital information such as heart rate, face expression, eye gazing as well as mouse movement and typing amount. In addition, we embed sensors into a mouse to monitor an environmental stress via temperature and humidity of user hand.

DOI： 10.1145/2968219.2971397

Language：English   Publishing type：Research paper (other academic)  

Smart Experiment Platform for Wearable Computing

Language：English   Publishing type：Research paper (other academic)  

Self-monitoring (tracking) of food and activity behavior is an evidence-based strategy for weight loss, and the main strategy employed in weight loss applications (apps). However, tracking is difficult and time-consuming and it is the main cause for abandoning usage of weight loss apps. We conducted a participatory design study with 26 participants to identify users' ideas on how to design better tracking experiences. Participants were interested in 2 main features: (i) adaptive interfaces, where the input methods and details required change based on context, and (ii) timely and context-Aware decision and motivation support.

DOI： 10.1145/296219.2968441

Language：English   Publishing type：Research paper (other academic)  

In this paper, we report our efforts to recognize human context through the things sensorized by our sensor platform As you know, many things wearing on your body have alread become a sensor to understand our activities. Mos famous things are watches and glasses. In the future, w suppose that unwearable things, such as cutlery, sport gea and instruments, would be also a sensor for recognizing ou activities. Compared with our movements, the movement o unwearable things will have more variation. Also, suitabl sensor and proper algorithm would be different according t the things. To enhance the senserization of various things we have proposed a platform, called SenStick, that consist of ultra tiny all-in-one sensor board and sophisticate mobile app. If using it, every researcher can try embeddin various sensor into various things and can focus on dat analysis. We explain the possibility of smart experimen platform with introducing various sensorized things on th platform.

DOI： 10.1145/2968219.2968268

Language：English   Publishing type：Research paper (other academic)  

In this demo, we present a novel rapid prototyping platform called SenStick2, which is all-in-one sensor with wireless charging. The main purpose of our platform is to sensorize our personal belongings easily and smartly for recognizing our living activities. The size of SenStick2 hardware (board) is 50mm(W) 10mm(H) 5mm(D) and its weight is around 3 (g) including a battery. On this tiny board, 8 sensors (acceleration, gyro, magnetic, light, UV, temperature, humidity, and pressure), flash memory, BLE, and battery are embedded in high density. The battery life in stand-Alone mode is more than 15 hours. For SenStick platform, we have developed the support software for iOS and Android OS. It can monitor the sensing data and also record the ground truth video simultaneously and synchronously. Furthermore, 3D CAD data of various case designs will be open for users in SenStick community site. By providing 3D case data, all the researcher can measure every tiny activities in a daily life easily and smartly.

DOI： 10.1145/2968219.2971399

Language：English   Publishing type：Research paper (other academic)  

In this paper, we propose a novel system called Downhill, which collects data for slope failure detection. Our system builds a Bluetooth Low Energy-based multi-hop communication network using smartphones. As a low-cost alternative to pile sensors, the system uses smartphones as sensor terminals to collect data. The smartphone sensors are placed in mountainous areas, which are prone to slope failures, and establish a MANET using their built-in BLE interface. If a smartphone detects ground movement, its collected acceleration data and GPS location data are sent to a server via the BLE network. These functions are executed by an application running on the smartphones. Hence, in the proposed system, it is possible to use the smartphone as the sensor terminal as long as one that has sufficient performance for use as the sensor terminal is prepared. To show the effectiveness of the system, we performed multi-hop communication experiments using Downhill in mountains that are at risk of landslides. Results show that Downhill is able to use effectively in the warning area of slope failure.

DOI： 10.1109/ICPPW.2016.18

Language：English   Publishing type：Research paper (other academic)  

EdgeComputing and Fog Computing are new paradigms where data processing is executed in or on the edge of networks to mitigate cloud server load. However, EdgeComputing and Fog Computing still need powerful servers on the edge of networks which impose additional costs for deployments. We proposed a platform called IFoT (Information Flow of Things) that efficiently performs distributed processing as well as distribution and analysis of data streams near their sources based on "Process On Our Own (PO3)" concept. In IFoT, processing of tasks for cloud servers is delegated to an ad-hoc distributed system consisting of proximity IoT devices for distributed real-time stream processing. In this demonstration, we show a face recognition system for person tracking developed on top of IFoT middleware which locally processes video streams in real-time and in a distributed manner by using computational resources of IoT devices.

DOI： 10.1109/ICDCS.2016.101

Language：Japanese