SoftPatch: Unsupervised Anomaly Detection with Noisy Data
This addresses a critical issue for real-world anomaly detection systems where clean data is often unavailable, though it is an incremental improvement over existing methods.
The paper tackled the problem of unsupervised anomaly detection with noisy training data, which limits practical applications, by proposing SoftPatch, a memory-based method that denoises data at the patch level and outperforms state-of-the-art methods on benchmarks like MVTecAD and BTAD in noisy scenes.
Although mainstream unsupervised anomaly detection (AD) algorithms perform well in academic datasets, their performance is limited in practical application due to the ideal experimental setting of clean training data. Training with noisy data is an inevitable problem in real-world anomaly detection but is seldom discussed. This paper considers label-level noise in image sensory anomaly detection for the first time. To solve this problem, we proposed a memory-based unsupervised AD method, SoftPatch, which efficiently denoises the data at the patch level. Noise discriminators are utilized to generate outlier scores for patch-level noise elimination before coreset construction. The scores are then stored in the memory bank to soften the anomaly detection boundary. Compared with existing methods, SoftPatch maintains a strong modeling ability of normal data and alleviates the overconfidence problem in coreset. Comprehensive experiments in various noise scenes demonstrate that SoftPatch outperforms the state-of-the-art AD methods on the MVTecAD and BTAD benchmarks and is comparable to those methods under the setting without noise.