Multi-Scale Patch-Based Representation Learning for Image Anomaly Detection and Segmentation.
Chin-Chia TsaiTsung-Hsuan WuShang-Hong LaiPublished in: WACV (2022)
Keyphrases
- multiscale
- anomaly detection
- image representation
- image segmentation
- segmentation algorithm
- connectionist systems
- edge detection
- segmentation method
- multiple scales
- intrusion detection
- image analysis
- anomalous behavior
- network intrusion detection
- input image
- learning process
- natural images
- test images
- network traffic
- unsupervised learning
- detecting anomalies
- behavior analysis
- intrusion detection system
- image features
- learning algorithm
- texture synthesis
- supervised learning
- face recognition
- cumulative sum