Domain-Adversarial Anatomical Graph Networks for Cross-User Human Activity Recognition

2025-05-08

AI Summary

Cross-user human activity recognition (HAR) suffers from poor generalization due to inter-subject variations in sensor placement, biomechanical dynamics, and behavioral patterns. Method: This paper proposes an anatomy-informed graph neural network framework that explicitly encodes biomechanically grounded anatomical relationships—interconnected, analogous, and lateral units—as a multi-relational enhanced graph structure. We further introduce a variational edge feature extractor and an adversarial domain generalization mechanism incorporating gradient reversal layers to learn user-invariant, biomechanically consistent representations. Contribution/Results: Evaluated on OPPORTUNITY and DSADS benchmarks, our approach achieves state-of-the-art performance, significantly improving robustness to unseen users and cross-domain generalization. It establishes a novel, transferable paradigm for HAR by unifying anatomical priors with geometric deep learning and domain-agnostic representation learning.

Technology Category

• Application Domains: Humanities & Computational Social Science
• Computer Vision: Video Understanding & Activity Analysis
• Machine Learning: Graph-based Machine Learning

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation

Abstract

Cross-user variability in Human Activity Recognition (HAR) remains a critical challenge due to differences in sensor placement, body dynamics, and behavioral patterns. Traditional methods often fail to capture biomechanical invariants that persist across users, limiting their generalization capability. We propose an Edge-Enhanced Graph-Based Adversarial Domain Generalization (EEG-ADG) framework that integrates anatomical correlation knowledge into a unified graph neural network (GNN) architecture. By modeling three biomechanically motivated relationships together-Interconnected Units, Analogous Units, and Lateral Units-our method encodes domain-invariant features while addressing user-specific variability through Variational Edge Feature Extractor. A Gradient Reversal Layer (GRL) enforces adversarial domain generalization, ensuring robustness to unseen users. Extensive experiments on OPPORTUNITY and DSADS datasets demonstrate state-of-the-art performance. Our work bridges biomechanical principles with graph-based adversarial learning by integrating information fusion techniques. This fusion of information underpins our unified and generalized model for cross-user HAR.

Problem

• Addressing cross-user variability in Human Activity Recognition
• Capturing biomechanical invariants across different users
• Enhancing generalization with adversarial domain adaptation

Innovation

• Graph neural network with anatomical correlation knowledge
• Variational Edge Feature Extractor for user variability
• Gradient Reversal Layer for domain generalization