CharacterGen: Efficient 3D Character Generation from Single Images with Multi-View Pose Canonicalization.
Hao-Yang PengJia-Peng ZhangMeng-Hao GuoYan-Pei CaoShi-Min HuPublished in: CoRR (2024)
Keyphrases
- camera views
- multi view
- multiple cameras
- multiple views
- dynamic scenes
- camera calibration
- d objects
- pose estimation
- single view
- three dimensional
- depth map
- view synthesis
- bundle adjustment
- photometric stereo
- image based rendering
- semi supervised
- scene reconstruction
- multi view images
- camera parameters
- multi views
- visual hull
- object recognition
- computer vision
- multi view clustering
- multi view learning
- viewpoint
- surface reconstruction
- multi view stereo
- face alignment
- camera pose
- geometric constraints
- point correspondences
- depth information
- pairwise
- input image
- face recognition
- range images
- test images
- free viewpoint
- image sequences