Source-free domain adaptation based on label reliability for cross-domain bearing fault diagnosis

2025-03-11

AI Summary

This work addresses cross-domain bearing fault diagnosis under source-free domain adaptation (SFDA). Unlike conventional SFDA approaches that rely solely on high-confidence pseudo-labels, we propose a full-sample utilization framework that systematically exploits informative cues from *unreliable* pseudo-labels—a novel paradigm in SFDA. Methodologically, we introduce a data-augmentation-driven multi-view label voting mechanism to assess pseudo-label reliability, jointly optimize feature- and label-space distributions, and incorporate maximum-entropy regularization to mitigate negative transfer. The framework dynamically balances discriminability and feature diversity. Extensive experiments on the CWRU and PU benchmarks demonstrate significant improvements over state-of-the-art SFDA methods. Our code is publicly available.

Technology Category

• Machine Learning: Transfer, Domain Adaptation, Multi-Task Learning
• Knowledge Representation and Reasoning: Diagnosis and Abductive Reasoning
• Intelligent Robots: State Estimation

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Web Mining and Content Analysis: Web data provenance, reliability, and authenticity
• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems

Abstract

Source-free domain adaptation (SFDA) has been exploited for cross-domain bearing fault diagnosis without access to source data. Current methods select partial target samples with reliable pseudo-labels for model adaptation, which is sub-optimal due to the ignored target samples. We argue that every target sample can contribute to model adaptation, and accordingly propose in this paper a novel SFDA-based approach for bearing fault diagnosis that exploits both reliable and unreliable pseudo-labels. We develop a data-augmentation-based label voting strategy to divide the target samples into reliable and unreliable ones. We propose to explore the underlying relation between feature space and label space by using the reliable pseudo-labels as ground-truth labels, meanwhile, alleviating negative transfer by maximizing the entropy of the unreliable pseudo-labels. The proposed method achieves well-balance between discriminability and diversity by taking advantage of reliable and unreliable pseudo-labels. Extensive experiments are conducted on two bearing fault benchmarks, demonstrating that our approach achieves significant performance improvements against existing SFDA-based bearing fault diagnosis methods. Our code is available at https://github.com/BdLab405/SDALR.

Problem

• Improves cross-domain bearing fault diagnosis without source data.
• Utilizes both reliable and unreliable pseudo-labels for model adaptation.
• Enhances performance by balancing discriminability and diversity in labels.

Innovation

• Utilizes reliable and unreliable pseudo-labels
• Implements data-augmentation-based label voting
• Maximizes entropy for unreliable pseudo-labels