2021 IEEE International Conference on Image Processing
19-22 September 2021 • Anchorage, Alaska, USA
Imaging Without Borders
| Paper ID | MLR-APPL-IP-8.5 | ||
| Paper Title | Unsupervised variability normalization for anomaly detection | ||
| Authors | Aitor Artola, ENS Paris-Saclay, Université Paris-Saclay, France; Yannis Kolodziej, Visionairy, France; Jean-Michel Morel, Thibaud Ehret, ENS Paris-Saclay, Université Paris-Saclay, France | ||
| Session | MLR-APPL-IP-8: Machine learning for image processing 8 | ||
| Location | Area E | ||
| Session Time: | Wednesday, 22 September, 14:30 - 16:00 | ||
| Presentation Time: | Wednesday, 22 September, 14:30 - 16:00 | ||
| Presentation | Poster | ||
| Topic | Applications of Machine Learning: Machine learning for image processing | ||
| IEEE Xplore Open Preview | Click here to view in IEEE Xplore | ||
| Abstract | Anomaly detectors are necessary to automatize industrial quality control. However, crafting such detectors is difficult due to the complexity and variability of the object even when working only with rigid objects. We show that adding a deep learning normalization step as a preprocessing step to model based detectors allows for better and more robust detections. This self-supervised normalization neural network is trained on non-anomalous data only. The proposed preprocessing method, followed by an automatic detector, achieves state-of-the-art results on rigid objects from the MvTec dataset. | ||