Pix4Point: Image Pretrained Transformers for 3D Point Cloud Understanding.
Guocheng QianXingdi ZhangAbdullah HamdiBernard GhanemPublished in: CoRR (2022)
Keyphrases
- single image
- image data
- input image
- multiscale
- image features
- template matching
- image representation
- image content
- image classification
- image collections
- image retrieval
- image pixels
- edge detection
- test images
- high resolution
- hough transform
- image segmentation
- image regions
- segmentation method
- feature points
- image analysis
- low level
- region of interest
- similarity measure
- false matches
- keypoints
- image matching
- image quality
- vector field
- pixel values
- computer vision
- grey level
- intensity profile
- virtual camera