Self-Supervised Pre-Training with Masked Shape Prediction for 3D Scene Understanding.
Li JiangZetong YangShaoshuai ShiVladislav GolyanikDengxin DaiBernt SchielePublished in: CVPR (2023)
Keyphrases
- d scene
- single image
- depth map
- image based rendering
- coordinate frame
- planar surfaces
- photorealistic
- scene understanding
- camera parameters
- optical flow
- rigidly moving
- indoor scenes
- scene geometry
- multi view
- camera viewpoint
- vector field
- data sets
- flow estimation
- computer vision
- camera positions
- view synthesis
- high quality
- training set
- viewpoint
- face recognition
- scene structure
- scene flow
- video sequences