ToCoAD: Two-Stage Contrastive Learning for Industrial Anomaly Detection

Unsupervised industrial anomaly detection methods can identify product defects in industrial images using only normal samples for training. Existing approaches perform well on public industrial image datasets but struggle with specific anomaly types due to the domain gap between pretrained feature extractors and target-specific domains. To tackle this issue, this article presents a two-stage training strategy, called ToCoAD. In the first stage, a discriminative network is trained by using synthetic anomalies in a self-supervised learning manner. This network is then utilized in the second stage to provide a negative feature guide, aiding in the training of the feature extractor through bootstrap contrastive learning. This approach enables the model to progressively learn the distribution of anomalies specific to industrial datasets, effectively enhancing its generalizability to various types of anomalies. Extensive experiments are conducted to demonstrate the effectiveness of our proposed two-stage training strategy, and our model produces competitive performance, achieving pixel-level AUROC scores of 98.21%, 98.43%, and 97.70% on MVTec AD, VisA, and BTAD, respectively.