AA-CLIP: Enhancing Zero-shot Anomaly Detection via Anomaly-Aware CLIP
This addresses the need for improved anomaly detection in applications like defect and lesion detection, though it is incremental as it builds on CLIP with a two-stage adaptation.
The paper tackles the problem of CLIP's limited discrimination between normal and abnormal features in zero-shot anomaly detection by proposing AA-CLIP, which enhances anomaly discrimination in text and visual spaces while preserving generalization, achieving state-of-the-art results in industrial and medical applications.
Anomaly detection (AD) identifies outliers for applications like defect and lesion detection. While CLIP shows promise for zero-shot AD tasks due to its strong generalization capabilities, its inherent Anomaly-Unawareness leads to limited discrimination between normal and abnormal features. To address this problem, we propose Anomaly-Aware CLIP (AA-CLIP), which enhances CLIP's anomaly discrimination ability in both text and visual spaces while preserving its generalization capability. AA-CLIP is achieved through a straightforward yet effective two-stage approach: it first creates anomaly-aware text anchors to differentiate normal and abnormal semantics clearly, then aligns patch-level visual features with these anchors for precise anomaly localization. This two-stage strategy, with the help of residual adapters, gradually adapts CLIP in a controlled manner, achieving effective AD while maintaining CLIP's class knowledge. Extensive experiments validate AA-CLIP as a resource-efficient solution for zero-shot AD tasks, achieving state-of-the-art results in industrial and medical applications. The code is available at https://github.com/Mwxinnn/AA-CLIP.