Coarse-to-Fine Volumetric Prediction for Single-Image 3D Human Pose.
Georgios PavlakosXiaowei ZhouKonstantinos G. DerpanisKostas DaniilidisPublished in: CVPR (2017)
Keyphrases
- single image
- coarse to fine
- multiscale
- multiresolution
- multiple images
- shape recovery
- hierarchical segmentation
- object detection
- d scene
- image registration
- super resolution
- image pairs
- light source
- intrinsic images
- shape from shading
- dynamic programming
- indoor scenes
- image segmentation
- body parts
- shape representation
- specular reflection
- image sequences
- three dimensional