This approach is named small-sample dense teacher assistant knowledge distillation (SS-DTAKD), which aims to enable bearing fault diagnosis with small samples and limited on-board storage resources.