Bridging Generalization and Personalization in Human Activity Recognition via On-Device Few-Shot Learning
This work addresses the problem of user-specific variations degrading HAR performance for wearable device users, offering an incremental improvement through efficient on-device adaptation.
The paper tackles the challenge of balancing generalization and personalization in Human Activity Recognition (HAR) by proposing an on-device few-shot learning framework that adapts to new users with minimal data, achieving accuracy improvements of 3.73%, 17.38%, and 3.70% on three benchmark datasets.
Human Activity Recognition (HAR) with different sensing modalities requires both strong generalization across diverse users and efficient personalization for individuals. However, conventional HAR models often fail to generalize when faced with user-specific variations, leading to degraded performance. To address this challenge, we propose a novel on-device few-shot learning framework that bridges generalization and personalization in HAR. Our method first trains a generalizable representation across users and then rapidly adapts to new users with only a few labeled samples, updating lightweight classifier layers directly on resource-constrained devices. This approach achieves robust on-device learning with minimal computation and memory cost, making it practical for real-world deployment. We implement our framework on the energy-efficient RISC-V GAP9 microcontroller and evaluate it on three benchmark datasets (RecGym, QVAR-Gesture, Ultrasound-Gesture). Across these scenarios, post-deployment adaptation improves accuracy by 3.73\%, 17.38\%, and 3.70\%, respectively. These results demonstrate that few-shot on-device learning enables scalable, user-aware, and energy-efficient wearable human activity recognition by seamlessly uniting generalization and personalization. The related framework is open sourced for further research\footnote{https://github.com/kangpx/onlineTiny2023}.