RevColV2: Exploring Disentangled Representations in Masked Image Modeling.
Qi HanYuxuan CaiXiangyu ZhangPublished in: NeurIPS (2023)
Keyphrases
- image features
- input image
- image data
- multiscale
- image content
- image representation
- single image
- image classification
- image retrieval
- pixel values
- template matching
- high resolution
- image analysis
- image pixels
- test images
- region of interest
- grey level
- vector field
- image segmentation
- hough transform
- keypoints
- mid level
- image collections
- image structure
- image descriptors
- shape properties