Co-attention CNNs for Unsupervised Object Co-segmentation.
Kuang-Jui HsuYen-Yu LinYung-Yu ChuangPublished in: IJCAI (2018)
Keyphrases
- object segmentation
- multiple objects
- fully unsupervised
- level set
- segmentation algorithm
- image segmentation
- text segmentation
- bayesian image segmentation
- pixel level
- segmentation accuracy
- image regions
- unsupervised learning
- focus of attention
- video objects
- semi supervised
- signed distance
- segmentation of natural images
- energy function
- object shape
- medical images
- image segments
- graph cut segmentation
- saliency map
- global shape
- bounding box
- moving objects
- complex objects
- d objects
- neural network
- visual attention
- object regions
- fully automatic
- cellular neural networks
- edge detection
- segmentation method
- computer vision
- image analysis
- segmentation of textured images
- supervised learning
- geometric shapes
- graph cuts
- object contours
- shape model
- test images
- background subtraction
- perceptual grouping
- texture segmentation
- shape prior
- object model
- partial occlusion