CorVS: Person Identification via Video Trajectory-Sensor Correspondence in a Real-World Warehouse
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866914124259131392 |
|---|---|
| author | Kano, Kazuma Mori, Yuki Katayama, Shin Urano, Kenta Yonezawa, Takuro Kawaguchi, Nobuo |
| author_facet | Kano, Kazuma Mori, Yuki Katayama, Shin Urano, Kenta Yonezawa, Takuro Kawaguchi, Nobuo |
| contents | 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. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2510_26369 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | CorVS: Person Identification via Video Trajectory-Sensor Correspondence in a Real-World Warehouse Kano, Kazuma Mori, Yuki Katayama, Shin Urano, Kenta Yonezawa, Takuro Kawaguchi, Nobuo Machine Learning Computer Vision and Pattern Recognition Robotics 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. |
| title | CorVS: Person Identification via Video Trajectory-Sensor Correspondence in a Real-World Warehouse |
| topic | Machine Learning Computer Vision and Pattern Recognition Robotics |
| url | https://arxiv.org/abs/2510.26369 |