研究者詳細 - 米澤　拓郎

論文 - 米澤　拓郎

分割表示 301 件中 1 - 20 件目／全件表示 >>

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

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

DIGITAL HEALTH 12 巻頁： 20552076261420289 2026年

記述言語：英語掲載種別：研究論文（学術雑誌）出版者・発行元：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

Representating Negative State of a Robot’s Abdominal/Thoracic Breathing Accompanied Movement System 査読有り

Fujii R., Nishiguchi Y., Moriyama K., Yonezawa T.

Lecture Notes in Computer Science 16337 LNCS 巻頁： 201 - 215 2026年

記述言語：日本語掲載種別：研究論文（学術雑誌）出版者・発行元：Lecture Notes in Computer Science

This paper presents an upper-body robot system that expresses involuntary abdominal and thoracic (chest) breathing. The system employs servomotors to generate breathing-related movements in both the shoulder and abdominal regions in response to the robot’s breathing parameters. By adjusting the speed, stability, and range of motion of each servomotor, the system aims to enhance nonverbal communication by expressing emotional and physical states such as excitement, anxiety, panic, fatigue, and debility. Experimental results suggest that combining multiple elements of breathing expression can improve the robot’s ability to evoke empathetic responses through emotionally expressive physical behavior.

DOI： 10.1007/978-3-032-12801-0_14

Multi-City Next Location Prediction through Mobility-Derived Multi-Pattern Transfer Learning 査読有り

Terashima H., Tamura N., Shoji K., Urano K., Yonezawa T., Kawaguchi N.

33rd ACM Sigspatial International Conference on Advances in Geographic Information Systems ACM Sigspatial 2025 頁： 1218 - 1221 2025年11月

記述言語：日本語掲載種別：研究論文（国際会議プロシーディングス）出版者・発行元：33rd ACM Sigspatial International Conference on Advances in Geographic Information Systems ACM Sigspatial 2025

In this study, we propose MoDeMIT (Mobility-Derived Multi-Insight Transfer), a human mobility prediction method that transfers various patterns shared across cities, extracted from mobility histories. Existing studies on human mobility prediction have primarily focused on learning the mobility patterns of users in the target city, without considering applications in cities with limited data. Moreover, relying solely on mobility patterns poses inherent limitations on prediction accuracy. The proposed MoDeMIT addresses these issues by defining and transferring multiple patterns shared across cities, such as lifestyle patterns and large-scale mobility patterns derived from mobility histories, as well as mobility patterns. This approach enables improvements in prediction accuracy compared to existing methods. We validate the effectiveness of MoDeMIT using real-world human mobility datasets. Furthermore, in the HuMob Challenge 2025 (GISCUP), MoDeMIT achieved a GEOBLEU score of [Average: 0.1632, CityA: 0.1504, CityB: 0.1471, CityC: 0.1801, CityD: 0.1753] and ranked within the top five teams.

DOI： 10.1145/3748636.3771311

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年2月

Efficient Edge AI based Annotation and Detection Framework for Logistics Warehouses 査読有り国際誌

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

IEEE Consumer Communications & Networking Conference 2025 2025年1月

Fixture-aware Panoramic Ceiling Image Synthesis Using an Omnidirectional Camera

Kohama D., Nagata Y., Yasutake K., Katayama S., Urano K., Yonezawa T., Kawaguchi N.

Journal of Information Processing 33 巻 ( 0 ) 頁： 245 - 258 2025年

記述言語：英語掲載種別：研究論文（学術雑誌）出版者・発行元：一般社団法人情報処理学会

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.

DOI： 10.2197/ipsjjip.33.245

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査読有り

住谷昌彦, 大竹祐子, 金岡晃, 青木崇行, 米澤拓郎

日本VR医学会学術大会プログラム・抄録集 2025 巻 ( 0 ) 頁： 10 2025年

Soccer Player Tracking Using Multiple Features for ID Switch Correction in Fixed-View Videos 査読有り

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

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

記述言語：日本語掲載種別：研究論文（学術雑誌）出版者・発行元：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

Smartphone-Based Activity Recognition in a LogisticsWarehouse Using Self-supervised Representation Learning 査読有り

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 頁： 1067 - 1073 2025年

記述言語：日本語掲載種別：研究論文（学術雑誌）出版者・発行元：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

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

Nagata, Y; Kohama, D; Watanabe, Y; Katayama, S; Urano, K; Yonezawa, T; Kawaguchi, N

IEEE ACCESS 13 巻頁： 114159 - 114178 2025年

記述言語：英語掲載種別：研究論文（学術雑誌）出版者・発行元：IEEE Access

The exponential growth of Real-World Data (RWD), primarily collected from IoT sensors and spanning domains such as mobility, environment, and energy consumption, presents critical challenges due to its scale, heterogeneity, and structural variability. Traditional compression methods often fail to adapt efficiently to these complexities, leading to sub-optimal storage and analysis performance. Additionally, while several metadata schemas exist to enhance the availability of RWD, smaller organizations often lack the resources to create and manage metadata effectively. This paper introduces RWD Profile, an automatically generatable metadata schema for RWD, and SemantiPack, which applies tailored compression to data fields in the individual RWD based on RWD Profile. RWD Profile has two kinds of metadata, Structural Profile and Semantic Profile. Structural Profile is generated through rule-based systems, while Semantic Profile is generated using large language models (LLMs) to capture semantic data properties. On the other hand, SemantiPack performs compression at the data field level in RWD using RWD Profile to achieve higher compression ratios. Experimental results demonstrate up to 23.2% improved compression rate for JSON and 19.3% for CSV compared to conventional methods while maintaining a faster or equivalent processing speed compared to conventional methods. Furthermore, SemantiPack supports lossy compression for applications prioritizing storage over precision. This research not only improves compression efficiency but also establishes a scalable solution for automated analysis and sustainable data utilization, paving the way for advancements in RWD management.

DOI： 10.1109/ACCESS.2025.3583829

Real-time Activity-Aware Video Streaming While Preserving Privacy 査読有り

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 頁： 236 - 240 2025年

記述言語：日本語掲載種別：研究論文（学術雑誌）出版者・発行元：Ubicomp Companion 2025 Companion of the 2025 ACM International Joint Conference on Pervasive and Ubiquitous Computing

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

Quantifying Work-Stress in Surgical Intensive Care Unit Nurses: A Multimodal Activity-Aware Physiological Analysis 査読有り 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 Part of 213533 巻頁： 127 - 133 2025年

記述言語：日本語掲載種別：研究論文（学術雑誌）出版者・発行元：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

Mitigating Latency Effects in VR180 Teleoperation via Virtual Robotic Arm Overlay 査読有り

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年

記述言語：日本語掲載種別：研究論文（学術雑誌）出版者・発行元：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

MetaUniTour: Enhancing Engagement and Intercultural Understanding through A Metaverse Tourism Platform 査読有り

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

Proceedings 2025 International Conference on Metaverse Computing Networking and Applications Metacom 2025 頁： 125 - 130 2025年

記述言語：日本語掲載種別：研究論文（学術雑誌）出版者・発行元：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

Life Pattern-Based Human Attribute Estimation Using Only GPS Data 査読有り Open Access

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

IEEE ACCESS 13 巻頁： 131803 - 131822 2025年

記述言語：日本語掲載種別：研究論文（学術雑誌）出版者・発行元：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

Feel the Pain: An Interpretable Multimodal Approach for Physiological Signal-Based Pain Detection 査読有り

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 頁： 100 - 108 2025年

記述言語：日本語掲載種別：研究論文（学術雑誌）出版者・発行元：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

Efficient Edge AI based Annotation and Detection Framework for Logistics Warehouses 査読有り

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

2025 IEEE 22ND CONSUMER COMMUNICATIONS & NETWORKING CONFERENCE, CCNC 頁： 1 - 4 2025年

記述言語：日本語掲載種別：研究論文（国際会議プロシーディングス）出版者・発行元：Proceedings IEEE Consumer Communications and Networking Conference Ccnc

As global logistics demand increases, improving the efficiency of warehouse operations has become critical. To achieve this, it's crucial to identify inefficient tasks and layouts by recognizing various warehouse conditions. To address these challenges, we've constructed a large-scale camera infrastructure to convert object positions and movements into data. However, transmitting all video data to the cloud results in significant data transmission and power consumption. Edge AI cameras analyze and extract video data locally, transmitting only essential information and significantly reducing them. Edge AI cameras require a low-computation, high-accuracy object detection model due to limited computational power and complex warehouse environments. Furthermore, the same object appears differently based on the camera's position and angle. Therefore, customizing models for each camera improves accuracy, but the annotation cost would be very high. In this study, we propose a method to perform part of the training data generation on the camera, reducing data transmission and improving annotation efficiency.

DOI： 10.1109/CCNC54725.2025.10976142

その他リンク： https://dblp.uni-trier.de/db/conf/ccnc/ccnc2025.html#MoriAHKUYK25

Discovering Emerging Urban Hotspots Using Satellite Imagery and Human Mobility Data 査読有り

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 15803 巻頁： 110 - 120 2025年

記述言語：日本語掲載種別：研究論文（学術雑誌）出版者・発行元：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

Digitization Methods for a Logistics Warehouse Towards Digital Twin-Driven Optimization 査読有り

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 頁： 186 - 193 2025年

記述言語：日本語掲載種別：研究論文（学術雑誌）出版者・発行元：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

CorVS: Person Identification via Video Trajectory-Sensor Correspondence in a Real-World Warehouse 査読有り 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年

記述言語：日本語掲載種別：研究論文（学術雑誌）出版者・発行元：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

　前のページ - 次のページ　

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16