Embedded Inter-Subject Variability in Adversarial Learning for Inertial Sensor-Based Human Activity Recognition

2025-08-31

International Workshop on Machine Learning for Signal Processing

AI Summary

This work addresses the challenge of poor cross-user generalization in inertial sensor-based human activity recognition (HAR) caused by inter-individual variability. To mitigate this issue, the authors propose a novel deep adversarial learning framework that explicitly models and embeds inter-subject differences within the adversarial training process, thereby learning user-invariant feature representations. By enhancing subject invariance, the method effectively reduces the influence of individual-specific characteristics in the learned features. Experimental evaluation on three widely used HAR datasets under leave-one-subject-out (LOSO) cross-validation demonstrates that the proposed approach significantly diminishes inter-subject discrepancies in the feature space and achieves superior recognition accuracy compared to existing state-of-the-art methods in cross-user scenarios.

Technology Category

• Machine Learning: Adversarial Learning & Robustness
• Computer Vision: Adversarial Attacks & Robustness
• Multiagent Systems: Adversarial Agents

Application Category

• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs

Abstract

This paper addresses the problem of Human Activity Recognition (HAR) using data from wearable inertial sensors. An important challenge in HAR is the model's generalization capabilities to new unseen individuals due to inter-subject variability, i.e., the same activity is performed differently by different individuals. To address this problem, we propose a novel deep adversarial framework that integrates the concept of inter-subject variability in the adversarial task, thereby encouraging subject-invariant feature representations and enhancing the classification performance in the HAR problem. Our approach outperforms previous methods in three well-established HAR datasets using a leave-one-subject-out (LOSO) cross-validation. Further results indicate that our proposed adversarial task effectively reduces inter-subject variability among different users in the feature space, and it outperforms adversarial tasks from previous works when integrated into our framework. Code: https://github.com/FranciscoCalatrava/EmbeddedSubjectVariability.git

Problem

• Human Activity Recognition
• Inter-Subject Variability
• Inertial Sensors
• Generalization
• Subject-Invariant

Innovation

• adversarial learning
• inter-subject variability
• subject-invariant representation
• inertial sensor-based HAR
• leave-one-subject-out