Unsupervised Anomaly Segmentation using Image-Semantic Cycle Translation.
Chenxin LiYunlong ZhangJiongcheng LiYue HuangXinghao DingPublished in: CoRR (2021)
Keyphrases
- segmentation method
- image segmentation
- segmentation algorithm
- image analysis
- grey level
- multiscale
- image regions
- test images
- pixel level
- image features
- edge detection
- image content
- image data
- homogeneous regions
- image pixels
- watershed algorithm
- segmentation accuracy
- input image
- segmented images
- image classification
- segmentation of natural images
- energy function
- optimal segmentation
- single image
- microscopic images
- level set
- energy functional
- perceptual grouping
- region segmentation
- intensity images
- image segments
- gray level images
- superpixels
- bounding box
- image representation
- image retrieval
- graph cuts
- semi supervised
- object segmentation
- low depth of field
- foreground background separation
- berkeley segmentation dataset
- optical flow
- low level
- brain mr images
- semantic segmentation
- intensity distribution
- foreground and background
- region growing
- image segmentation algorithms
- microscopy images
- markov random field
- segmented regions
- segmentation errors
- region merging
- medical images
- gray value
- watershed transformation
- unsupervised learning
- pixel values