Patch-wise Auto-Encoder for Visual Anomaly Detection
This addresses anomaly detection in industrial applications without prior knowledge of anomalies, representing an incremental improvement over traditional auto-encoders.
The paper tackles the problem of unsupervised visual anomaly detection by proposing a patch-wise auto-encoder framework that enhances reconstruction ability for anomalies, achieving state-of-the-art performance on the Mvtec AD benchmark.
Anomaly detection without priors of the anomalies is challenging. In the field of unsupervised anomaly detection, traditional auto-encoder (AE) tends to fail based on the assumption that by training only on normal images, the model will not be able to reconstruct abnormal images correctly. On the contrary, we propose a novel patch-wise auto-encoder (Patch AE) framework, which aims at enhancing the reconstruction ability of AE to anomalies instead of weakening it. Each patch of image is reconstructed by corresponding spatially distributed feature vector of the learned feature representation, i.e., patch-wise reconstruction, which ensures anomaly-sensitivity of AE. Our method is simple and efficient. It advances the state-of-the-art performances on Mvtec AD benchmark, which proves the effectiveness of our model. It shows great potential in practical industrial application scenarios.