GenHAR: Generalizing Cross-domain Human Activity Recognition for Last-mile Delivery
Name
3770854.3783921.pdf
Size
3.18 MB
Format
Adobe PDF
Checksum (MD5)
695a0b1f94f9bf1ecc070f20ef639cfb
Author(s)
Hong, Zhiqing
Li, Zelong
Fan, Xiubin
Yang, Guang
Guo, Baoshen
Wang, Haotian
He, Tian
Zhang, Desheng
Date Issued
April 20, 2026
Publisher
Association for Computing Machinery
Citation
Zhiqing Hong, Zelong Li, Xiubin Fan, Guang Yang, Baoshen Guo, Haotian Wang, Tian He, and Desheng Zhang. 2026. GenHAR: Generalizing Cross-domain Human Activity Recognition for Last-mile Delivery. In Proceedings of the 32nd ACM SIGKDD Conference on Knowledge Discovery and Data Mining V.1 (KDD '26). Association for Computing Machinery, New York, NY, USA, 2208–2219.
Version
Final published version
Abstract
Human Activity Recognition (HAR) has shown remarkable effectiveness in various applications, such as smart healthcare and intelligent manufacturing. However, a major challenge faced by HAR is the distribution shift across different sensor data domains, which often leads to decreased performance when deployed for real-world applications. To address this issue, this paper introduces GenHAR, a novel framework designed to mitigate the domain gap by learning domain-invariant sensor representations. GenHAR aims to enhance the generalization capabilities of HAR on target domains purely with data from the source domain. The key novelty of GenHAR lies in two aspects. Firstly, GenHAR tokenizes sensor data and learns correlations among frequency sensor channel dimensions to improve the robustness of HAR models. Secondly, GenHAR improves the efficiency via selective masking and an efficient attention mechanism. We conduct a systematic analysis of GenHAR by comparing it with state-of-the-art HAR methods on real-world human activity datasets. Results show that GenHAR outperforms state-of-the-art methods by 9.97% in accuracy, and reduces Floating Point Operations by 6.4 times. Moreover, we deploy GenHAR at a leading logistics company in 4 cities, and have detected 2.15 billion real-time activities. We release our code at: https://github.com/Sensor-Foundation-Model/GenHAR.
Description
KDD ’26, Jeju Island, Republic of Korea
MIT Department
Singapore-MIT Alliance in Research and Technology (SMART)
Terms of Use
Creative Commons Attribution
Persistent DSpace Link
DOI of Published Version
https://doi.org/10.1145/3770854.3783921
