Unsupervised anomaly segmentation via deep feature reconstruction.
Yong ShiJie YangZhiquan QiPublished in: Neurocomputing (2021)
Keyphrases
- point distribution
- text segmentation
- segmentation algorithm
- bayesian image segmentation
- color and texture information
- anomaly detection
- segmentation method
- intrusion detection
- level set
- image segmentation
- segmentation of natural images
- fully unsupervised
- image reconstruction
- image analysis
- feature vectors
- object segmentation
- preprocessing step
- supervised learning
- unsupervised learning
- high resolution
- segmented regions
- shape prior
- image processing
- medical images
- segmentation of textured images
- multiscale
- deep learning
- normalized cut
- medical imaging
- semi supervised
- geometric features
- selection scheme
- image regions
- moving objects
- positron emission tomography
- discrete tomography
- image features
- d objects
- face recognition
- three dimensional