“…However, the domain adaptation method in transfer learning performs poorly in the actual transfer fault diagnosis scenario, and the adaptive enhancement of fault features cannot be achieved during the transfer process. In order to address above issues, Xu et al [18] proposed a novel approach called the Siamese Feature Discriminative Attention Enhanced Transfer (SDAET) fault diagnosis method. This method adds the attention-enhanced network in the feature extraction stage, and can adaptively enhance the features in the network learning process and improve the accuracy of diagnosis.…”