Unsupervised anomaly instance segmentation for baggage threat recognition.
Taimur HassanSamet AkçayMohammed BennamounSalman KhanNaoufel WerghiPublished in: J. Ambient Intell. Humaniz. Comput. (2023)
Keyphrases
- unsupervised learning
- recognition rate
- correct recognition
- numeral strings
- text segmentation
- recognition accuracy
- image parsing
- fully unsupervised
- image segmentation
- segmentation algorithm
- anomaly detection
- level set
- symbol recognition
- segmentation method
- bayesian image segmentation
- action recognition
- pattern recognition
- character segmentation
- x ray
- segmentation of natural images
- automatic recognition
- medical images
- recognition algorithm
- object segmentation
- fully automatic
- object recognition
- features extraction
- skeleton extraction
- feature extraction and matching
- multiscale
- countermeasures
- character recognition
- shape prior
- region growing
- semantic segmentation
- semi supervised
- object detection
- segmentation accuracy
- activity recognition
- network traffic
- energy function
- computer vision
- supervised learning