Multi-level Contrast Network for Wearables-based Joint Activity Segmentation and Recognition

TL;DR

This paper introduces a multi-level contrastive learning approach combined with a multi-stage temporal convolutional network to improve joint activity segmentation and recognition in wearable-based human activity recognition, addressing multi-class window issues.

Contribution

It proposes a novel multi-level contrastive loss integrated with MS-TCN for enhanced activity segmentation and recognition, a new approach in wearable HAR.

Findings

Significant improvement in recognition accuracy on two public datasets.

Effective handling of multi-class window problem in HAR.

Enhanced robustness against inter-class similarity and intra-class heterogeneity.

Abstract

Human activity recognition (HAR) with wearables is promising research that can be widely adopted in many smart healthcare applications. In recent years, the deep learning-based HAR models have achieved impressive recognition performance. However, most HAR algorithms are susceptible to the multi-class windows problem that is essential yet rarely exploited. In this paper, we propose to relieve this challenging problem by introducing the segmentation technology into HAR, yielding joint activity segmentation and recognition. Especially, we introduce the Multi-Stage Temporal Convolutional Network (MS-TCN) architecture for sample-level activity prediction to joint segment and recognize the activity sequence. Furthermore, to enhance the robustness of HAR against the inter-class similarity and intra-class heterogeneity, a multi-level contrastive loss, containing the sample-level and segment-level contrast, has been proposed to learn a well-structured embedding space for better activity segmentation and recognition performance. Finally, with comprehensive experiments, we verify the effectiveness of the proposed method on two public HAR datasets, achieving significant improvements in the various evaluation metrics.