Image2Point: 3D Point-Cloud Understanding with 2D Image Pretrained Models.
Chenfeng XuShijia YangTomer GalantiBichen WuXiangyu YueBohan ZhaiWei ZhanPeter VajdaKurt KeutzerMasayoshi TomizukaPublished in: ECCV (37) (2022)
Keyphrases
- input image
- multiscale
- image content
- image data
- single image
- image features
- template matching
- image retrieval
- image classification
- image segmentation
- region of interest
- bayesian framework
- low level
- image collections
- image regions
- image analysis
- edge detection
- image representation
- high resolution
- image pixels
- random fields
- grey level
- false matches
- test images
- vector field
- statistical model
- segmentation method
- image registration
- motion estimation
- multiresolution
- video sequences
- similarity measure