Unsupervised Multi-View Visual Anomaly Detection via Progressive Homography-Guided Alignment
Xintao Chen, Xiaohao Xu, Bozhong Zheng, Yun Liu, Yingna Wu
Abstract
Unsupervised visual anomaly detection from multi-view images presents a significant challenge: distinguishing genuine defects from benign appearance variations caused by viewpoint changes. Existing methods, often designed for single-view inputs, treat multiple views as a disconnected set of images, leading to inconsistent feature representations and a high false-positive rate. To address this, we introduce ViewSense-AD (VSAD), a novel framework that learns viewpoint-invariant representations by explicitly modeling geometric consistency across views. At its core is our Multi-View Alignment Module (MVAM), which leverages homography to project and align corresponding feature regions between neighboring views. We integrate MVAM into a View-Align Latent Diffusion Model (VALDM), enabling progressive and multi-stage alignment during the denoising process. This allows the model to build a coherent and holistic understanding of the object
BibTeX
@inproceedings{aaai2026_unsupervisedmult,
title = {Unsupervised Multi-View Visual Anomaly Detection via Progressive Homography-Guided Alignment},
author = {Xintao Chen and Xiaohao Xu and Bozhong Zheng and Yun Liu and Yingna Wu},
booktitle = {AAAI 2026},
year = {2026}
}