Unleashing Foundation Models via Tri-Memory Bank for Structural and Logical Anomaly Detection

Overview

To unify logical and structural anomaly detection, we propose three complementary strategies:

1. Retain patch-level image features to construct a patch-level image memory bank $M_P$, thereby preserving the model’s capability to detect structural anomalies.
2. Introduce a logic-aware textual description of images to significantly enhance logical anomalies detection of unified anomaly detection frameworks by constructing a class-level text memory bank $M_T$.
3. Introduce object-level image features from segmented images to build an object-level image memory bank $M_O$, which compensates for the inability of patch-level features to fully capture the shape and structure of individual objects.