A Weight-aware-based Multi-source Unsupervised Domain Adaptation Method for Human Motion Intention Recognition

2024-04-19

arXiv.org

AI Summary

Unsupervised domain adaptation (UDA) for cross-subject human motion intention (HMI) recognition in exoskeleton robots suffers from degraded generalization due to unaddressed distribution heterogeneity across multiple source subjects. Method: We propose the first multi-source UDA framework with adaptive source weighting, extending marginal discrepancy distance (MDD) to the multi-source setting and deriving a theoretical upper bound on target-domain generalization error. Our approach integrates a lightweight network architecture, adversarial feature generation, an ensemble classifier-based min-max game, learnable source-domain weighting, and joint multi-source MDD optimization. Results: On HMI recognition tasks, our method significantly outperforms state-of-the-art UDA approaches, achieving higher classification accuracy while maintaining real-time inference capability. Both theoretical analysis and empirical evaluation consistently validate the effectiveness and soundness of the proposed framework.

Technology Category

• Machine Learning: Transfer, Domain Adaptation, Multi-Task Learning
• Multiagent Systems: Adversarial Agents
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Accurate recognition of human motion intention (HMI) is beneficial for exoskeleton robots to improve the wearing comfort level and achieve natural human-robot interaction. A classifier trained on labeled source subjects (domains) performs poorly on unlabeled target subject since the difference in individual motor characteristics. The unsupervised domain adaptation (UDA) method has become an effective way to this problem. However, the labeled data are collected from multiple source subjects that might be different not only from the target subject but also from each other. The current UDA methods for HMI recognition ignore the difference between each source subject, which reduces the classification accuracy. Therefore, this paper considers the differences between source subjects and develops a novel theory and algorithm for UDA to recognize HMI, where the margin disparity discrepancy (MDD) is extended to multi-source UDA theory and a novel weight-aware-based multi-source UDA algorithm (WMDD) is proposed. The source domain weight, which can be adjusted adaptively by the MDD between each source subject and target subject, is incorporated into UDA to measure the differences between source subjects. The developed multi-source UDA theory is theoretical and the generalization error on target subject is guaranteed. The theory can be transformed into an optimization problem for UDA, successfully bridging the gap between theory and algorithm. Moreover, a lightweight network is employed to guarantee the real-time of classification and the adversarial learning between feature generator and ensemble classifiers is utilized to further improve the generalization ability. The extensive experiments verify theoretical analysis and show that WMDD outperforms previous UDA methods on HMI recognition tasks.

Problem

• Recognizing human motion intention despite individual motor differences
• Addressing multi-source domain variation in unsupervised adaptation
• Improving exoskeleton robot interaction via adaptive weight-aware UDA

Innovation

• Weight-aware multi-source UDA algorithm (WMDD)
• Extended margin disparity discrepancy (MDD) theory
• Lightweight network with adversarial learning