Two-stage Co-segmentation Network Based on Discriminative Representation for Recovering Human Mesh from Videos.
Boyang ZhangKehua MaSuping WuZhixiang YuanPublished in: CVPR (2023)
Keyphrases
- perceptual information
- parametric representation
- scene representation
- video sequences
- temporal segmentation
- region growing
- image segmentation
- segmentation algorithm
- level set
- edge detection
- human activities
- shape prior
- video clips
- video data
- multiscale
- volumetric images
- hierarchical representation
- object shape
- video frames
- segmentation method
- d mesh
- fully unsupervised
- medical images
- geometric shapes
- video content
- foreground background segmentation
- feature extraction