Unsupervised industrial image ensemble anomaly detection based on object pseudo-anomaly generation and normal image feature combination enhancement.
Haoyuan ShenBaolei WeiYizhong MaXiaoyu GuPublished in: Comput. Ind. Eng. (2023)
Keyphrases
- anomaly detection
- image features
- image content
- intrusion detection
- unsupervised anomaly detection
- detecting anomalies
- anomalous behavior
- one class support vector machines
- image representation
- unsupervised learning
- network intrusion detection
- input image
- image regions
- object features
- intrusion detection system
- network traffic
- normal behavior
- image segmentation
- detecting anomalous
- negative selection algorithm
- image retrieval
- network anomaly detection
- object classes
- object recognition
- neural network
- misuse detection
- behavior analysis
- network security
- scale space
- detect anomalies
- similarity measure
- remote sensing
- abnormal behavior
- text mining
- probabilistic model
- feature vectors
- pairwise
- feature selection
- computer vision
- data sets