Gated Adaptation for Continual Learning in Human Activity Recognition

2026-03-08

AI Summary

This work addresses catastrophic forgetting in continual learning for human activity recognition on wearable devices by proposing a parameter-efficient, channel-level gated modulation framework. The approach enables cross-user incremental learning without uploading sensitive user data, replay buffers, or task-specific regularization. By freezing the pre-trained backbone network and adapting only through diagonal scaling of feature channels—requiring fewer than 2% trainable parameters—the method effectively mitigates representational drift while balancing model stability and plasticity. Evaluated on the PAMAP2 dataset across eight sequential users, the proposed framework reduces the forgetting rate from 39.7% to 16.2% and achieves a final accuracy of 77.7%, substantially outperforming standard continual learning baselines.

Technology Category

• Machine Learning: Life-Long and Continual Learning
• Search and Optimization: Learning to Search
• Natural Language Processing: Learning & Optimization for NLP

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: Machine-in-the-loop, human agency and autonomy

Abstract

Wearable sensors in Internet of Things (IoT) ecosystems increasingly support applications such as remote health monitoring, elderly care, and smart home automation, all of which rely on robust human activity recognition (HAR). Continual learning systems must balance plasticity (learning new tasks) with stability (retaining prior knowledge), yet AI models often exhibit catastrophic forgetting, where learning new tasks degrades performance on earlier ones. This challenge is especially acute in domain-incremental HAR, where on-device models must adapt to new subjects with distinct movement patterns while maintaining accuracy on prior subjects without transmitting sensitive data to the cloud. We propose a parameter-efficient continual learning framework based on channel-wise gated modulation of frozen pretrained representations. Our key insight is that adaptation should operate through feature selection rather than feature generation: by restricting learned transformations to diagonal scaling of existing features, we preserve the geometry of pretrained representations while enabling subject-specific modulation. We provide a theoretical analysis showing that gating implements a bounded diagonal operator that limits representational drift compared to unconstrained linear transformations. Empirically, freezing the backbone substantially reduces forgetting, and lightweight gates restore lost adaptation capacity, achieving stability and plasticity simultaneously. On PAMAP2 with 8 sequential subjects, our approach reduces forgetting from 39.7% to 16.2% and improves final accuracy from 56.7% to 77.7%, while training less than 2% of parameters. Our method matches or exceeds standard continual learning baselines without replay buffers or task-specific regularization, confirming that structured diagonal operators are effective and efficient under distribution shift.

Problem

• Continual Learning
• Human Activity Recognition
• Catastrophic Forgetting
• Domain-Incremental Learning
• Wearable Sensors

Innovation

• gated adaptation
• continual learning
• human activity recognition
• catastrophic forgetting
• parameter-efficient learning