CLIP2Scene: Towards Label-efficient 3D Scene Understanding by CLIP.
Runnan ChenYouquan LiuLingdong KongXinge ZhuYuexin MaYikang LiYuenan HouYu QiaoWenping WangPublished in: CVPR (2023)
Keyphrases
- d scene
- single image
- depth map
- scene understanding
- image based rendering
- photorealistic
- planar surfaces
- scene structure
- indoor scenes
- scene geometry
- optical flow
- moving camera
- camera viewpoint
- camera positions
- coordinate frame
- ego motion
- camera parameters
- depth information
- view synthesis
- view dependent
- image sequences
- three dimensional
- scene flow
- data sets
- high resolution
- feature space
- scene representation
- planar patches