RiEMann: Near Real-Time SE(3)-Equivariant Robot Manipulation without Point Cloud Segmentation.
Chongkai GaoZhengrong XueShuying DengTianhai LiangSiqi YangLin ShaoHuazhe XuPublished in: CoRR (2024)
Keyphrases
- point cloud
- real time
- vision system
- surface reconstruction
- point sets
- mobile robot
- manipulation tasks
- segmentation algorithm
- structure from motion
- laser scanner
- dense reconstruction
- segmentation method
- point cloud data
- shape prior
- image segmentation
- level set
- medical images
- stereo camera
- iterative closest point
- plane fitting
- image regions
- single image
- graph cuts
- multi view
- d objects
- image processing
- machine learning