Review
Open access
2026
Integrated Transfer and Self-Supervised Learning for Data-Efficient Bearing Fault Diagnosis: A Survey Toward Industry 5.0
The analysis indicates that existing studies frequently investigate TL and SSL methods for bearing fault diagnosis independently, rely predominantly on controlled benchmark datasets, and provide limited validation of cross-machine and cross-condition generalization.
Faiza Ayyob, Shumaila Javaid, Shuai-Yong Li et al.
· IEEE Access · 0 citations