Learning Disentangled Behaviour Patterns for Wearable-based Human Activity Recognition

TL;DR

This paper introduces a Behaviour Pattern Disentanglement framework that isolates activity signals from noise in wearable-based human activity recognition, improving feature learning despite variability and limited data.

Contribution

The novel BPD framework effectively disentangles activity patterns from irrelevant noise using adversarial training, enhancing existing deep learning models for HAR.

Findings

Improved accuracy on four public HAR datasets.

Demonstrated flexibility of BPD with various deep learning models.

Effective noise disentanglement in complex activity signals.

Abstract

In wearable-based human activity recognition (HAR) research, one of the major challenges is the large intra-class variability problem. The collected activity signal is often, if not always, coupled with noises or bias caused by personal, environmental, or other factors, making it difficult to learn effective features for HAR tasks, especially when with inadequate data. To address this issue, in this work, we proposed a Behaviour Pattern Disentanglement (BPD) framework, which can disentangle the behavior patterns from the irrelevant noises such as personal styles or environmental noises, etc. Based on a disentanglement network, we designed several loss functions and used an adversarial training strategy for optimization, which can disentangle activity signals from the irrelevant noises with the least dependency (between them) in the feature space. Our BPD framework is flexible, and it can be used on top of existing deep learning (DL) approaches for feature refinement. Extensive experiments were conducted on four public HAR datasets, and the promising results of our proposed BPD scheme suggest its flexibility and effectiveness. This is an open-source project, and the code can be found at http://github.com/Jie-su/BPD