OrganicHAR: Towards Activity Discovery in Organic Settings for Privacy Preserving Sensors Using Efficient Video Analysis
Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies · 02.12.2025 · 9 (4) · 1-32
DOI: 10.1145/3770674 ↗Kurz gefasst
Fachartikel in der Zeitschrift Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies, online erschienen am 02.12.2025. Verfasst von Prasoon Patidar und 11 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 82 Literaturangaben. Gefördert von: National Science Foundation.
Abstract (Originalsprache)
Deploying human activity recognition (HAR) at home is still rare because sensor signals vary wildly across houses, people, and time, essentially requiring in-situ data collection and training. Prior approaches use cameras to generate training labels for privacy-preserving sensors (LiDAR, RADAR, Thermal), but this forces sensors to detect predefined activities that cameras can see yet the sensors themselves cannot reliably distinguish. In this work, we introduce OrganicHAR, an activity discovery framework that inverts this relationship by placing sensor capabilities at the center of activity discovery. Our approach identifies naturally occurring signal patterns using privacy-preserving sensors, leverages Vision Language Models (VLMs) only during these key moments for scene understanding, and discovers discrete activity labels at granularities that these sensors can reliably detect. Our evaluation with 12 participants demonstrates OrganicHAR's effectiveness: it achieves 79% accuracy for coarse (4-5) activities using only basic ambient sensors (radar, lidar, thermal arrays), and 73% accuracy for fine-grained (8-9) activities when a wearable IMU, depth, and pose sensor are added. OrganicHAR maintains 77% accuracy on average across configurations while discovering 4-8 categories per user (15 across all users) tailored to each environment and sensor capabilities. By triggering video processing only at key moments identified by local sensors, we reduce queries to VLM by 90%, enabling practical and privacy-preserving activity recognition in natural settings.
Förderung
- National Science Foundation
Zitieren
APA
Patidar, P., Arakawa, R., Graça, R., Moutinho, R., Soares, A., Vasconcelos, A., Talami, F., Silva, J. C. D., Silva, I., Santos, C. M., Goel, M., & Agarwal, Y. (2025). OrganicHAR: Towards Activity Discovery in Organic Settings for Privacy Preserving Sensors Using Efficient Video Analysis. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies, 9(4), 1-32. https://doi.org/10.1145/3770674
Vancouver
Patidar P, Arakawa R, Graça R, Moutinho R, Soares A, Vasconcelos A, et al. OrganicHAR: Towards Activity Discovery in Organic Settings for Privacy Preserving Sensors Using Efficient Video Analysis. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies. 2025;9(4):1-32. doi:10.1145/3770674
BibTeX
@article{Patidar2025,
author = {Patidar, Prasoon and Arakawa, Riku and Graça, Ricardo and Moutinho, Rúben and Soares, Adriano and Vasconcelos, Ana and Talami, Filippo and Silva, Joana Couto da and Silva, Inês and Santos, Cristina Mendes and Goel, Mayank and Agarwal, Yuvraj},
title = {OrganicHAR: Towards Activity Discovery in Organic Settings for Privacy Preserving Sensors Using Efficient Video Analysis},
journal = {Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies},
year = {2025},
volume = {9},
number = {4},
pages = {1--32},
doi = {10.1145/3770674}
}
EUVIMED