FlexUOD: The Answer to Real-world Unsupervised Image Outlier Detection

How many outliers are within an unlabeled and contaminated dataset? Despite a series of unsupervised outlier detection (UOD) approaches have been proposed, they cannot correctly answer this critical question, resulting in their performance instability across various real-world (varying contamination factor) scenarios. To address this problem, we propose FlexUOD, with a novel contamination factor estimation perspective. FlexUOD not only achieves its remarkable robustness but also is a general and plug-andplay framework, which can significantly improve the performance of existing UOD methods. Extensive experiments demonstrate that FlexUOD achieves state-of-the-art results as well as high efficacy on diverse evaluation benchmarks. Inlier Outlier Frequency (a) Low-contamination Factor Estimation (ground truth: 0.05) REGR KARCH Estimated Results REGR: 0.294 KARCH: 0.152 Ours: 0.057 Inlier Outlier Outlier Score Frequency (b) High-contamination Factor Estimation (ground truth: 0.3) KARCH REGR Estimated Results KARCH: 0.159 REGR: 0.124 Ours: 0.291