Anomaly Detection with Test Time Augmentation and Consistency Evaluation
This addresses the vulnerability of neural networks to unseen data for applications requiring reliable anomaly detection, though it is incremental as it builds on existing augmentation and consistency ideas.
The paper tackles the problem of deep neural networks assigning high confidence to out-of-distribution data by proposing TTA-AD, a post-hoc anomaly detection method that uses test-time augmentation and consistency evaluation, achieving comparable or better performance with a 60% to 90% reduction in running time compared to existing methods.
Deep neural networks are known to be vulnerable to unseen data: they may wrongly assign high confidence stcores to out-distribuion samples. Recent works try to solve the problem using representation learning methods and specific metrics. In this paper, we propose a simple, yet effective post-hoc anomaly detection algorithm named Test Time Augmentation Anomaly Detection (TTA-AD), inspired by a novel observation. Specifically, we observe that in-distribution data enjoy more consistent predictions for its original and augmented versions on a trained network than out-distribution data, which separates in-distribution and out-distribution samples. Experiments on various high-resolution image benchmark datasets demonstrate that TTA-AD achieves comparable or better detection performance under dataset-vs-dataset anomaly detection settings with a 60%~90\% running time reduction of existing classifier-based algorithms. We provide empirical verification that the key to TTA-AD lies in the remaining classes between augmented features, which has long been partially ignored by previous works. Additionally, we use RUNS as a surrogate to analyze our algorithm theoretically.