Revisiting Reverse Distillation for Anomaly Detection

Anomaly detection is an important application in largescale industrial manufacturing. Recent methods for this task have demonstrated excellent accuracy but come with a latency trade-off. Memory based approaches with dominant performances like PatchCore or Coupled-hyperspherebased Feature Adaptation (CFA) require an external memory bank, which significantly lengthens the execution time. Another approach that employs Reversed Distillation (RD) can perform well while maintaining low latency. In this paper, we revisit this idea to improve its performance, establishing a new state-of-the-art benchmark on the challenging MVTec dataset for both anomaly detection and localization. The proposed method, called RD++, runs six times faster than PatchCore, and two times faster than CFA but introduces a negligible latency compared to RD. We also experiment on the BTAD and Retinal OCT datasets to demonstrate our method's generalizability and conduct important ablation experiments to provide insights into its configurations. Source code will be available at https : / / github . com / tientrandinh / Revisiting-Reverse-Distillation. Recent research in anomaly detection, such as Patch-Core [27], and CFA [20], have achieved state-of-the-art performance in detecting and localizing anomalies. However, This CVPR paper is the Open Access version, provided by the Computer Vision Foundation. Except for this watermark, it is identical to the accepted version; the final published version of the proceedings is available on IEEE Xplore.