Class-independent sequential full image segmentation, using a convolutional net that finds a segment within an attention region, given a pointer pixel within this segment.
Sagi EppelPublished in: CoRR (2019)
Keyphrases
- probabilistic model
- homogeneous regions
- image segmentation
- segmented regions
- accurate segmentation
- segmentation algorithm
- segmentation result
- region growing
- segmentation scheme
- optimal segmentation
- input image
- gray level
- region segmentation
- coherent regions
- figure ground
- watershed segmentation
- region boundaries
- deformable models
- level set
- multiscale
- gradient magnitude
- level set method
- fully unsupervised
- computer vision
- visual attention
- graph cuts
- image processing