2 papers
cs.CV2026
XMatchAD: A Cross-Modal Matching Perspective on Reconstruction-based Anomaly Detection
Mingxiu Cai, Zhe Zhang, Gaochang Wu +1
The remarkable success of reconstruction-based methods in Unsupervised Anomaly Detection (UAD) lies in their ability to identify and localize anomalies by modeling discrepancies be…
cs.CV2025
Unified Unsupervised Anomaly Detection via Matching Cost Filtering
Zhe Zhang, Mingxiu Cai, Gaochang Wu +5
Unsupervised anomaly detection (UAD) aims to identify image- and pixel-level anomalies using only normal training data, with wide applications such as industrial inspection and med…