About this document

32433-Article Text-36501-1-2-20250410 by guosphia0329 is a document available to read on EtoBox.

The document presents LogicAD, a novel algorithm for explainable anomaly detection (AD) that leverages autoregressive multimodal Vision Language Models (AVLMs) for logical anomaly detection without the need for extensive visual annotations. LogicAD achieves state-of-the-art performance on public benchmarks, significantly outperforming existing methods in both anomaly detection accuracy and explainability. The approach integrates text feature extraction, format embedding, and a logic reasoner to effectively

Author

guosphia0329

Language

EN