Open-Vocabulary Point-Cloud Object Detection without 3D Annotation.
Yuheng LuChenfeng XuXiaobao WeiXiaodong XieMasayoshi TomizukaKurt KeutzerShanghang ZhangPublished in: CoRR (2023)
Keyphrases
- point cloud
- object detection
- surface reconstruction
- structure from motion
- computer vision
- laser scanner
- scene understanding
- multi class
- metadata
- point sets
- active learning
- triangular mesh
- object categories
- object classification
- stereo camera
- real world objects
- iterative closest point
- object recognition
- image regions
- image retrieval
- controlled vocabulary
- image sequences
- point cloud data