"Unsupervised, Online and On-The-Fly Anomaly Detection For Non-Stationary Image Distributions"
Declan GD McIntosh*, Alexandra Branzan Albu
Abstract
"We propose Online-InReaCh, the first fully unsupervised online method for detecting and localizing anomalies on-the-fly in image sequences while following non-stationary distributions. Previous anomaly detection methods are limited to supervised one-class classification or are unsupervised but still pre-compute their nominal model. Online-InReaCh can operate online by dynamically maintaining a nominal model of commonly occurring patches that associate well across image realizations of the underlying nominal distribution while removing stale previously nominal patches. Online-InReaCh, while competitive in previous offline benchmarks, also achieves 0.936 and 0.961 image- and pixel-wise AUROC when tested online on MVTecAD, where 23.8% of all randomly sampled images contain anomalies. Online-InReaCh’s performance did not correlate with anomaly proportion even to 33.5%. We also show that Online-InReaCh can integrate new nominal structures and distinguish anomalies after a single frame, even in the worst-case distribution shift from one training class to a new previously unseen testing class. Code: https://github.com/DeclanMcIntosh/Online_InReaCh"
BibTeX
@inproceedings{eccv2024_unsupervisedonli,
title = {"Unsupervised, Online and On-The-Fly Anomaly Detection For Non-Stationary Image Distributions"},
author = {Declan GD McIntosh* and Alexandra Branzan Albu},
booktitle = {ECCV 2024},
year = {2024}
}