Shared Latent Membership Enables Joint Shape Abstraction and Segmentation With Deformable Superquadrics.
Jiaxin LiHongxing WangJiawei TanJunsong YuanPublished in: IEEE Trans. Image Process. (2024)
Keyphrases
- shape prior
- deformable models
- deformable shapes
- level set
- segmentation algorithm
- shape model
- robust segmentation
- segmentation method
- image segmentation
- region based segmentation
- deformable template
- object segmentation
- multiple scales
- complex shapes
- energy functional
- shape analysis
- object shape
- shape variations
- geometric shapes
- medical images
- skeleton extraction
- active shape
- shape descriptors
- statistical shape model
- perceptual grouping
- variational framework
- image segments
- remote sensed images
- image gradient
- edge detection
- shape matching
- object boundaries
- prostate segmentation
- appearance cues
- shape decomposition
- prior shape
- prior information
- region growing
- global shape
- brain structures
- intensity distribution
- shape representation
- object detection
- multiscale
- image analysis
- image regions
- latent variables
- deformable surface
- shape features