VoxFormer: Sparse Voxel Transformer for Camera-based 3D Semantic Scene Completion.
Yiming LiZhiding YuChristopher B. ChoyChaowei XiaoJose M. AlvarezSanja FidlerChen FengAnima AnandkumarPublished in: CoRR (2023)
Keyphrases
- real scenes
- camera images
- scene geometry
- scene structure
- multiple images
- autocalibration
- fish eye
- three dimensional
- d scene
- ground plane
- uncalibrated cameras
- acquired images
- stereo camera
- moving camera
- imaging process
- active illumination
- image capture
- video rate
- ego motion
- video images
- video sequences
- defocused images
- line features
- single image
- camera tracking
- camera positions
- camera motion
- light conditions
- field of view
- vanishing points
- virtual camera
- camera calibration
- object motion
- multi camera
- image correspondences
- live video
- camera parameters
- camera views
- response function
- dynamic scenes
- optical axis
- point features
- vision system
- hand held
- white balance
- relative position
- single shot
- multiple cameras
- video camera
- moving platform
- viewing direction
- input image
- structure from motion
- single camera
- stereo vision
- moving objects
- optical flow
- pose estimation
- depth estimation
- camera movement
- multiple views
- epipolar geometry
- computer vision