Bridging Generalization and Personalization in Wearable Human Activity Recognition via On-Device Few-Shot Learning

2025-08-21

AI Summary

To address poor cross-user generalization and high post-deployment personalization overhead in wearable human activity recognition (HAR), this paper proposes a hybrid framework: “cross-user generalization first, followed by on-device few-shot adaptation.” The core innovation lies in updating only a lightweight classifier layer directly on resource-constrained edge devices—specifically, a RISC-V-based GAP9 microcontroller—enabling low-overhead, robust online few-shot learning. This marks the first realization of concept-drift-driven user adaptation on embedded platforms. The method integrates few-shot learning, embedded model compression, and user-specific fine-tuning. Experiments across three real-world deployment scenarios demonstrate accuracy improvements of 3.73%, 17.38%, and 3.70%, respectively, significantly enhancing both cross-user generalization capability and on-device adaptive efficiency.

Technology Category

• Machine Learning: Learning on the Edge & Model Compression
• Humans and AI: Interaction Techniques and Devices
• Cognitive Modeling & Cognitive Systems: Adaptive Behavior

Application Category

• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Responsible Web: Human-perceived consequences of algorithmic deployment on the web

Abstract

Human Activity Recognition (HAR) using wearable devices has advanced significantly in recent years, yet its generalization remains limited when models are deployed to new users. This degradation in performance is primarily due to user-induced concept drift (UICD), highlighting the importance of efficient personalization. In this paper, we present a hybrid framework that first generalizes across users and then rapidly adapts to individual users using few-shot learning directly on-device. By updating only the classifier layer with user-specific data, our method achieves robust personalization with minimal computational and memory overhead. We implement this framework on the energy-efficient RISC-V-based GAP9 microcontroller and validate it across three diverse HAR scenarios: RecGym, QVAR-Gesture, and Ultrasound-Gesture. Post-deployment adaptation yields consistent accuracy improvements of 3.73%, 17.38%, and 3.70% respectively. These results confirm that fast, lightweight, and effective personalization is feasible on embedded platforms, paving the way for scalable and user-aware HAR systems in the wild footnote{https://github.com/kangpx/onlineTiny2023}.

Problem

• Addressing user-induced concept drift in wearable activity recognition
• Enabling on-device personalization with minimal computational overhead
• Achieving robust adaptation across diverse human activity scenarios

Innovation

• On-device few-shot learning for personalization
• Updating only classifier layer with user data
• Implemented on RISC-V-based microcontroller for efficiency