Faculty Profiles - YONEZAWA Takuro

Papers - YONEZAWA Takuro

Division display 1 - 20 of about 307 ／ All the affair displays >>

Outfitting the quest for spatial spread of infections: A review of mobility datasets for population health modelling Reviewed Open Access

Tillayeva, N; Tamura, N; Urano, K; Yonezawa, T; Kawaguchi, N; Okumura, T

INFECTIOUS DISEASE MODELLING Vol. 11 ( 3 ) page： 1248 - 1267 2026.9

Language：English Publishing type：Research paper (scientific journal) Publisher：Infectious Disease Modelling

Understanding the spatial dynamics of infectious disease spread is essential for modeling population health. A key component in such modeling is human mobility data, which informs how infections propagate across time and space. This review provides a comprehensive survey of both real-world and synthetic mobility datasets that have been used in the context of infectious disease modeling. Through the survey, we identified 57 publicly available datasets—52 real-world and 5 synthetic—offering a structured overview of current data sources. Additionally, because real-world data are often inaccessible due to privacy or technical constraints, we provide a concise overview of pseudo-mobility data generation methodologies to contextualize the synthetic datasets and guide future data-production efforts. The review highlights the need to advance synthetic data generation methodologies and improve the accessibility of high-resolution mobility data for future research in this domain.

DOI： 10.1016/j.idm.2026.03.006

Optical flow visual stimuli induce body sway and visually-induced motion sickness in virtual reality system: A preliminary study Reviewed Open Access

Otake, Y; Kanaoka, A; Aoki, S; Osumi, M; Yonezawa, T; Sumitani, M

DIGITAL HEALTH Vol. 12 page： 20552076261420289 2026

Language：English Publishing type：Research paper (scientific journal) Publisher：Digital Health

Objectives: Visually-induced motion sickness (VIMS) remains an unsolved issue when using head-mounted displays (HMD) in immersive virtual reality (VR) systems. We investigate the influence of restricted visual field on sensorimotor incongruence of postural control in an immersive VR environment, as a surrogate marker of VIMS while standing steadily and viewing ambiguous optical-flow visual stimuli. Methods: Twenty-seven healthy participants wore HMD and viewed optical flow visual stimuli (called the plaid motion pattern) using a VR system. Visual stimuli were presented on a full-screen display, either only in the center or periphery of the visual field, or nowhere (Natural). To evaluate standing posture stability, we measured spatial (length and area of the center of the pressure (COP)) and temporal dynamics of body sway using a stabilometer. Subjective feelings of VIMS were assessed for each visual stimulus condition. Results: The full-screen condition significantly worsened the COP measurements and feelings. The COP areas of standing balance under the center of the visual field condition were significantly smaller than those under the full-screen condition; however, the periphery of the visual field condition was comparable to the full-screen condition. The condition effects for the conditions (natural, center of the visual field, periphery of the visual field and full-screen) were observed in the COP measurements and subjective feelings. Conclusion: Optical-flow visual stimuli can induce body sway, suggesting sensorimotor incongruence in postural control. Ancillary findings demonstrated that subsequent to full-screen presentation, the periphery of the visual field potentially contributes to spatial and temporal dynamics of body sway and the sensorimotor incongruence rather than the central area.

DOI： 10.1177/20552076261420289

VR180 遠隔ロボットアームシステムにおける操作支援のための情報提示手法の検討 Open Access

筒井秀斗, 渡邉企章, 宇佐美龍斗, 具志祐希, 渡辺圭貴, 林田望海, 志村魁哉, 浦野健太, 米澤拓郎, 河口信夫

第85回UBI研究会 2025.2

Efficient Edge AI based Annotation and Detection Framework for Logistics Warehouses Reviewed International journal

Yuki Mori, Yusuke Asai, Keisuke Higashiura, Shin Katayama, Kenta Urano, Takuro Yonezawa, Nobuo Kawaguchi

IEEE Consumer Communications & Networking Conference 2025 2025.1

Fixture-aware Panoramic Ceiling Image Synthesis Using an Omnidirectional Camera Open Access

Kohama Daiki, Yasutake Kazushige, Nagata Yoshiteru, Katayama Shin, Urano Kenta, Yonezawa Takuro, Kawaguchi Nobuo

Journal of Information Processing Vol. 33 ( 0 ) page： 245 - 258 2025

Language：English Publishing type：Research paper (scientific journal) Publisher：Information Processing Society of Japan

<p>The manual creation of a ceiling plan consumes many human resources to confirm the current state of existing buildings for renovation. Identifying the positions and the types of existing fixtures is crucial for drawing a ceiling plan. Therefore, to assist in drawing a ceiling plan, an efficient method that generates a reliable photorealistic whole ceiling surface image is essential. We present a method of fixture-aware panoramic ceiling image synthesis using an omnidirectional camera. A fixture-aware panoramic ceiling image is a photorealistic image that depicts the whole ceiling surface and no seams on ceiling fixtures for its reliability regarding assistance in drawing a ceiling plan. The proposed method is remarkably effortless because it requires the users to work only 2 procedures for synthesizing the image, 1) shooting an omnidirectional video underneath the ceiling and 2) fixture measurement in the image. In the experiment, we evaluated the resulting image for the positional and shape accuracy of the ceiling fixtures and the time efficiency of the measurement. Our proposed method demonstrated a positional accuracy of 2.25 times higher, a shape accuracy of 5.87 times higher, and a time efficiency of the measurement 11.09 times higher than the baseline.</p>

DOI： 10.2197/ipsjjip.33.245

Open Access

Scopus

CiNii Research

オプティカルフロー視覚刺激は、バーチャルリアリティシステムにおいて無意識の身体動揺と視覚誘発性気分不良を誘発する Reviewed Open Access

住谷昌彦, 金岡晃, 大竹祐子, 米澤拓郎, 青木崇行

日本VR医学会学術大会プログラム・抄録集 Vol. 2025 ( 0 ) page： 10 2025

Soccer Player Tracking Using Multiple Features for ID Switch Correction in Fixed-View Videos Reviewed

Mori, Y; Takagi, S; Katayama, S; Urano, K; Yonezawa, T; Kawaguchi, N

8TH INTERNATIONAL ACM WORKSHOP ON MULTIMEDIA CONTENT ANALYSIS IN SPORTS, MMSPORTS 2025 page： 189 - 195 2025

Language：Japanese Publishing type：Research paper (scientific journal) Publisher：Mmsports 2025 8th International ACM Workshop on Multimedia Content Analysis in Sports

Player tracking in soccer is becoming more important for tactical analysis and performance evaluation. Many tracking methods have been developed using broadcast videos. However, such videos are often unavailable for amateur matches, such as those in universities or high schools. Fixed-viewpoint cameras are easier to install, but they cannot move or zoom. Because of this, many players appear small in each frame, which makes it hard to detect, identify, and track them. Tracking becomes more difficult when players are close to each other. This often causes ID switches (mistakes in identifying players) and lowers accuracy. To solve this problem, we joined the SoccerTrack Challenge 2025, which focuses on tracking players in videos from fixed-viewpoint cameras. We developed a new method to improve tracking accuracy in this setting. In our method, we divided each frame into multiple regions and detected players in each region. This helped us detect even small players more accurately. We also used features like uniform color, jersey numbers, and goalkeeper identification to correct ID switches and improve tracking performance. As a result, our method achieved a HOTA score of 0.59 and ranked among the top teams in the challenge.

DOI： 10.1145/3728423.3759417

Web of Science

Scopus

Smartphone-Based Activity Recognition in a LogisticsWarehouse Using Self-supervised Representation Learning Reviewed

Watanabe, K; Kano, K; Hossain, T; Katayama, S; Urano, K; Yonezawa, T; Kawaguchi, N

COMPANION OF THE 2025 ACM INTERNATIONAL JOINT CONFERENCE ON PERVASIVE AND UBIQUITOUS COMPUTING, UBICOMP COMPANION 2025 page： 1067 - 1073 2025

Language：Japanese Publishing type：Research paper (scientific journal) Publisher：Ubicomp Companion 2025 Companion of the 2025 ACM International Joint Conference on Pervasive and Ubiquitous Computing

To address labor shortages in logistics, digital twins that digitize physical environments are gaining attention, as they enable visualization and simulation-based optimization of warehouse operations. However, digitizing worker activities is particularly challenging due to operational burden and difficulty in adapting to diverse tasks. In response to this issue, we propose a deep learning-based framework for automatic warehouse task recognition using smartphone inertial sensors (accelerometer and gyroscope). Our model combines a CNN pretrained on unlabeled data using SimCLR, a self-supervised representation learning method, to improve generalization, and a Transformer encoder to capture task-level context. We collected data in a logistics warehouse and conducted cross-validation using 6 hours of annotated data from 10 individuals. As a result, the proposed method achieved a weighted F1 score of 0.89, improving by 7 points over the model without pretraining.

DOI： 10.1145/3714394.3756198

Web of Science

Scopus

SemantiPack: An Efficient Real-World Data Compressor Using Structural and Semantic Metadata Reviewed Open Access

Yoshiteru Nagata, Daiki Kohama, Yoshiki Watanabe, Shin Katayama, Kenta Urano, Takuro Yonezawa, Nobuo Kawaguchi

IEEE ACCESS Vol. 13 page： 114159 - 114178 2025

Real-time Activity-Aware Video Streaming While Preserving Privacy Reviewed

Shimura, K; Kano, K; Hossain, T; Katayama, S; Urano, K; Taguchi, S; Deguchi, H; Sakai, H; Yonezawa, T; Kawaguchi, N

COMPANION OF THE 2025 ACM INTERNATIONAL JOINT CONFERENCE ON PERVASIVE AND UBIQUITOUS COMPUTING, UBICOMP COMPANION 2025 page： 236 - 240 2025

We present ActStream, a remote communication system that selectively shares the user and objects relevant to their current activities. By combining vision-language models and real-time segmentation, ActStream detects and highlights only activity-relevant elements, enhancing awareness while preserving privacy. The system identifies key objects through user focus or interaction, synthesizes them with the user's image, and streams the result to a remote partner. ActStream aims to provide useful context without exposing unnecessary information, highlighting the potential of selective sharing in privacy-sensitive remote collaboration.

DOI： 10.1145/3714394.3754399

Web of Science

Scopus

Quantifying Work-Stress in Surgical Intensive Care Unit Nurses: A Multimodal Activity-Aware Physiological Analysis Reviewed Open Access

Hossain, T; Rahaman, MS; Yamashita, K; Kono, A; Guowei, M; Oyama, S; Yonezawa, T

PROCEEDINGS OF THE 2025 ACM INTERNATIONAL SYMPOSIUM ON WEARABLE COMPUTERS, ISWC 2025 Vol. Part of 213533 page： 127 - 133 2025

Language：Japanese Publishing type：Research paper (scientific journal) Publisher：Proceedings International Symposium on Wearable Computers Iswc

Work-related stress in critical care settings threatens both nurse well-being and patient safety. While understanding and accurately quantifying this stress is important in these high-stakes environments, it remains a significant challenge due to the inherent complexity of clinical workflows and the wide variability in individual stress responses. To address this, our study investigates how physiological signals captured via wearable sensors correspond to nurses' perceived stress during Surgical Intensive Care Unit (SICU) activities. We captured physiological indicators of stress using two wearable devices: a wrist-worn EmbracePlus and an upper arm-worn Apple Watch as nurses engaged in a range of simulated SICU activities. These physiological signals were then synchronized with perceived stress, captured using validated self-report questionnaires. Our analysis shows significant variations in stress responses across different SICU activities, with elevated heart rate variability (HRV) and increased perceived stress observed during physically demanding activities such as ''Patient Transfer''. A machine learning model trained on combined physiological and contextual features related to nursing activities achieved 80% accuracy under leave-one-participant-out cross-validation, demonstrating the potential of wearable sensing for activity-aware stress recognition. Principal Component Analysis (PCA) further showed that the wrist-worn sensor produced more structured and separable HRV patterns than the arm-worn device, suggesting the critical role of sensor placement and device characteristics in reliable stress assessment.

DOI： 10.1145/3715071.3750406

Open Access

Web of Science

Scopus

Multi-City Next Location Prediction through Mobility-Derived Multi-Pattern Transfer Learning Reviewed

Haru Terashima, Naoki Tamura, Kazuyuki Shoji, Kenta Urano, Takuro Yonezawa, Nobuo Kawaguchi

Proceedings of the 33rd ACM International Conference on Advances in Geographic Information Systems page： 1218 - 1221 2025

Mitigating Latency Effects in VR180 Teleoperation via Virtual Robotic Arm Overlay Reviewed

Tsutsui S., Watanabe K., Usami R., Gushi Y., Watanabe Y., Shimura K., Hayashida N., Jimenez F., Urano K., Yonezawa T., Kawaguchi N.

15th International Conference on Mobile Computing and Ubiquitous Networking Icmu 2025 2025

Language：Japanese Publishing type：Research paper (scientific journal) Publisher：15th International Conference on Mobile Computing and Ubiquitous Networking Icmu 2025

Many studies have investigated remote robotic arm operation systems using virtual reality environments. However, conventional VR teleoperation systems suffer from communication latency, which makes it difficult for operators to receive immediate visual feedback of the robot's posture. This mismatch between input and visual feedback may increase cognitive and psychological load and degrade task performance. We propose a teleoperation system that overlays a virtual robotic arm to provide immediate visual feedback of user operation prior to receiving the delayed video from the physical robotic arm. By superimposing a virtual robotic arm on the stereo video, the system enables operators to simultaneously maintain immersion in the remote environment and perceive the input posture of the robotic arm. We evaluated information presentation methods through remote robotic arm task experiments in VR and user feedback. The results confirmed that, under high-latency conditions, the virtual robotic arm display reduced operator anxiety and helped maintain confidence during teleoperation.

DOI： 10.23919/ICMU65253.2025.11219125

Scopus

MetaUniTour: Enhancing Engagement and Intercultural Understanding through A Metaverse Tourism Platform Reviewed

Matsumoto, K; Watanabe, Y; Shimura, K; Hayashida, N; Katayama, S; Urano, K; Kawaguchi, N; Aoki, S; Leleito, E; Yonezawa, T

2025 INTERNATIONAL CONFERENCE ON METAVERSE COMPUTING, NETWORKING AND APPLICATIONS, METACOM page： 125 - 130 2025

Language：Japanese Publishing type：Research paper (scientific journal) Publisher：Proceedings 2025 International Conference on Metaverse Computing Networking and Applications Metacom 2025

We propose "MetaUniTour", a remote tourism platform designed to address key limitations in conventional Metaverse-based tourism, including perceptual and experiential asymmetries between tourists and hosts, the lack of collaborative engagement due to individualized exploration and high barriers to participation. MetaUniTour is structured around three core design principles: (1) spatial design that creates "mixed spaces"integrating cultural elements from both tourists and hosts; (2) experience design based on semi-structured, group-oriented activities that balance individual autonomy and collective participation; and (3) a browser-based, device-independent system architecture that ensures broad accessibility. A two-month field deployment, "Reality of Kenya,"virtually connected Higashiyama Zoo in Nagoya, Japan, with Kenya's Maasai Mara and attracted 3,615 participants. A follow-up survey answered by 66 people revealed that over 60% perceived spatial integration between the two locations, also participants found the experience enjoyable and felt a sense of unity through experienced intercultural exchanges. Our findings demonstrate that MetaUniTour offers an effective design framework for fostering symmetrical cultural exchange and immersive shared experiences with any devices, suggesting promising potential for future applications in tourism, education, and cross-cultural communication.

DOI： 10.1109/MetaCom65502.2025.00028

Web of Science

Scopus

Life Pattern-Based Human Attribute Estimation Using Only GPS Data Reviewed Open Access

Shoji, K; Terashima, H; Tamura, N; Katayama, S; Urano, K; Yonezawa, T; Kawaguchi, N

IEEE ACCESS Vol. 13 page： 131803 - 131822 2025

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

Analysis of large-scale GPS data enables the summarization and structuring of human mobility characteristics, providing insights into residents' behavior patterns. However, existing methods analyze sequences of individual-derived stay labels, such as home or workplace, which do not accurately reflect actual area functions and can fail to distinguish individuals based on differences in workplaces. This limitation arises because annotating stay locations and purposes in GPS data is challenging due to the complexity of urban structures and constraints on data quality. In this study, we propose ArLPH, a framework that takes only GPS data as input, models mobility from the perspective of 'when and what types of places he/she visited,' and enables the interpretation of life patterns. ArLPH consists of two modules: 'Area Modeling,' which represents urban areas based on usage patterns, and 'Human Modeling,' which represents individuals based on behavior patterns. By the processing flow of area modeling to human modeling, it becomes possible to analyze human mobility while considering area characteristics. Through an evaluation of a large-scale, real-world GPS dataset collected from tens of thousands of smartphone users, we demonstrate that ArLPH enables the characterization of individuals based on their life patterns and estimates attributes such as occupation (e.g., office workers or homemakers) and lifestyle (e.g., those who work frequently on weekends).

DOI： 10.1109/ACCESS.2025.3591811

Open Access

Web of Science

Scopus

Feel the Pain: An Interpretable Multimodal Approach for Physiological Signal-Based Pain Detection Reviewed

Tazin, T; Hossain, T; Yamashita, K; Kinoshita, T; Tag, B; Oyama, S; Yonezawa, T; Salim, F

COMPANION PUBLICATION OF THE 27TH INTERNATIONAL CONFERENCE ON MULTIMODAL INTERACTION, ICMI 2025 COMPANION page： 100 - 108 2025

Language：Japanese Publishing type：Research paper (scientific journal) Publisher：Icmi 2025 Companion Publication of the 27th International Conference on Multimodal Interaction

Pain assessment is essential in clinical settings, but remains largely subjective, relying on verbal self-reports that are impractical for non-communicative patients. In this study, we propose an interpretable multimodal approach for automatic pain detection using physiological signals. Leveraging 65 participants (23 female, 42 male) multimodal data from the AI4Pain 2025 challenge, we systematically explore the contributions of multiple sensor modalities, Electrodermal Activity (EDA), Blood Volume Pulse (BVP), Respiration (RESP), and Peripheral Oxygen Saturation (SpO<inf>2</inf>) to classify pain into No Pain, Low Pain, and High Pain categories. After careful pre-processing and feature extraction, XGBoost outperformed other machine learning and deep learning models. In particular, the combination of only EDA and BVP features achieved 80% validation accuracy. In addition, the model using combined EDA and BVP features achieved high classification accuracy across all pain levels, particularly for High Pain (AUC = 0.92) and No Pain (AUC = 0.96), demonstrating strong discriminative ability. Using all modalities, the XGBoost model achieved a test accuracy of 60.92%, marking an 8.62% improvement over the multimodal test baseline. Notably, our model outperformed the baseline results across all modalities on both the validation and test sets. Our model also incorporates HRV features derived from BVP and EDA emphasizes interpretability through analysis of permutation feature importance. The results demonstrate that multimodal physiological data, when processed effectively, enable accurate and explainable pain detection, a step toward real-time, deployable pain monitoring systems in healthcare.

DOI： 10.1145/3747327.3764786

Web of Science

Scopus

Efficient Edge AI Based Annotation and Detection Framework for Logistics Warehouses. Reviewed

Yuki Mori, Yusuke Asai, Keisuke Higashiura, Shin Katayama, Kenta Urano, Takuro Yonezawa, Nobuo Kawaguchi

CCNC page： 1 - 4 2025

Discovering Emerging Urban Hotspots Using Satellite Imagery and Human Mobility Data Reviewed

Tillayeva, N; Terashima, H; Tamura, N; Shoji, K; Hossain, T; Katayama, S; Urano, K; Yonezawa, T; Kawaguchi, N

DISTRIBUTED, AMBIENT AND PERVASIVE INTERACTIONS, DAPI 2025, PT II Vol. 15803 page： 110 - 120 2025

Language：Japanese Publishing type：Research paper (scientific journal) Publisher：Lecture Notes in Computer Science

Identifying and prioritizing newly emerging points of interest in urban areas is critical for enhancing citizens’ exploration and engagement with their cities. However, traditional discovery channels—such as local media or social networks—often fail to capture real-time urban transformations comprehensively. In this paper, we propose a data-driven methodology that integrates satellite-based change detection with human mobility analytics to reveal, rank, and recommend newly formed or transformed urban hotspots. Specifically, we employ Synthetic Aperture Radar (SAR) Rapid and EAsy Change detection in radar TIme-series by Variation coefficient (REACTIV) to detect structural changes in satellite imagery, and we combine these insights with Area2Vec embeddings derived from BlogWatcher data (2019–2024) to measure shifts in human movement patterns. Our experiments on the rapidly evolving city of Nagoya demonstrate a strong correlation between physical urban transformations and mobility changes, enabling us to pinpoint and prioritize locations with the highest potential appeal. This approach has broad applications for city residents, tourists, businesses, and urban planners by facilitating more informed and timely exploration of dynamic urban spaces.

DOI： 10.1007/978-3-031-92980-9_7

Web of Science

Scopus

Digitization Methods for a Logistics Warehouse Towards Digital Twin-Driven Optimization Reviewed

Kawaguchi, N; Asai, Y; Kano, K; Takaki, K; Mori, Y; Suzuki, Y; Watanabe, K; Gushi, Y; Katayama, S; Urano, K; Yonezawa, T; Hashiguchi, S

2025 IEEE INTERNATIONAL CONFERENCE ON SMART COMPUTING, SMARTCOMP page： 186 - 193 2025

Language：Japanese Publishing type：Research paper (scientific journal) Publisher：Proceedings 2025 IEEE International Conference on Smart Computing Smartcomp 2025

The rapid growth of e-commerce, rising consumer expectations, and labor shortages in Japan pose challenges for logistics warehouse optimization. While automation has advanced outbound operations, inbound processes remain inefficient. This study presents a comprehensive approach to digitizing and optimizing large-scale logistics warehouses, with a case study at TRUSCO NAKAYAMA Corp. in partnership with Nagoya University. Key technologies include a large-scale camera array, multi-camera object tracking, and smartphone-based task estimation. Our contributions include real-time tracking of personnel and packages, cooperative annotation for improved object recognition, and synthetic data augmentation. Additionally, truck berth analysis and indoor localization enhance operational efficiency. To optimize worker shifts and warehouse layout, we apply Factorization Machine Quantum Annealing (FMQA), achieving a 37.4% reduction in lead times and a 14.3% decrease in labor hours. A visualization tool enables warehouse operators to make data-driven decisions. This research demonstrates the potential of digital transformation in logistics and provides a scalable framework for broader industry adoption.

DOI： 10.1109/SMARTCOMP65954.2025.00035

Web of Science

Scopus

CorVS: Person Identification via Video Trajectory-Sensor Correspondence in a Real-World Warehouse Reviewed Open Access

Kano K., Mori Y., Katayama S., Urano K., Yonezawa T., Kawaguchi N.

Proceedings of the 15th International Conference on Indoor Positioning and Indoor Navigation Ipin 2025 2025

Language：Japanese Publishing type：Research paper (scientific journal) Publisher：Proceedings of the 15th International Conference on Indoor Positioning and Indoor Navigation Ipin 2025

Worker location data is key to higher productivity in industrial sites. Cameras are a promising tool for localization in logistics warehouses since they also offer valuable environmental contexts such as package status. However, identifying individuals with only visual data is often impractical. Accordingly, several prior studies identified people in videos by comparing their trajectories and wearable sensor measurements. While this approach has advantages such as independence from appearance, the existing methods may break down under real-world conditions. To overcome this challenge, we propose CorVS, a novel data-driven person identification method based on correspondence between visual tracking trajectories and sensor measurements. Firstly, our deep learning model predicts correspondence probabilities and reliabilities for every pair of a trajectory and sensor measurements. Secondly, our algorithm matches the trajectories and sensor measurements over time using the predicted probabilities and reliabilities. We developed a dataset with actual warehouse operations and demonstrated the method's effectiveness for real-world applications.

DOI： 10.1109/IPIN66788.2025.11213222

Scopus

　 PREV - NEXT 　

<YONEZAWA Takuro>

Papers - YONEZAWA Takuro