Subject Invariant Contrastive Learning for Human Activity Recognition

TL;DR

This paper proposes Subject-Invariant Contrastive Learning (SICL), a new loss function that enhances human activity recognition models' ability to generalize across different subjects by suppressing subject-specific cues.

Contribution

The paper introduces SICL, a novel loss function that improves subject-invariant feature learning in contrastive learning for HAR, demonstrating significant performance gains.

Findings

SICL improves HAR accuracy by up to 11% over traditional methods.

SICL effectively suppresses subject-specific cues in sensor signals.

The loss function is adaptable across various self-supervised and supervised frameworks.

Abstract

The high cost of annotating data makes self-supervised approaches, such as contrastive learning methods, appealing for Human Activity Recognition (HAR). Effective contrastive learning relies on selecting informative positive and negative samples. However, HAR sensor signals are subject to significant domain shifts caused by subject variability. These domain shifts hinder model generalization to unseen subjects by embedding subject-specific variations rather than activity-specific features. As a result, human activity recognition models trained with contrastive learning often struggle to generalize to new subjects. We introduce Subject-Invariant Contrastive Learning (SICL), a simple yet effective loss function to improve generalization in human activity recognition. SICL re-weights negative pairs drawn from the same subject to suppress subject-specific cues and emphasize activity-specific information. We evaluate our loss function on three public benchmarks: UTD-MHAD, MMAct, and DARai. We show that SICL improves performance by up to 11% over traditional contrastive learning methods. Additionally, we demonstrate the adaptability of our loss function across various settings, including multiple self-supervised methods, multimodal scenarios, and supervised learning frameworks.