PETR: Position Embedding Transformation for Multi-view 3D Object Detection.
Yingfei LiuTiancai WangXiangyu ZhangJian SunPublished in: ECCV (27) (2022)
Keyphrases
- multi view
- d objects
- multiple views
- single view
- object recognition
- viewpoint
- multiple cameras
- depth map
- object detection
- three dimensional
- pose estimation
- camera calibration
- photometric stereo
- range data
- camera parameters
- detection method
- shape descriptors
- bundle adjustment
- visual hull
- d mesh
- texture mapping
- human body
- computer vision
- input image
- object classes
- surface reconstruction
- range images