On Efficient Transformer-Based Image Pre-training for Low-Level Vision.
Wenbo LiXin LuShengju QianJiangbo LuPublished in: IJCAI (2023)
Keyphrases
- low level vision
- image understanding
- image data
- energy minimization
- image restoration
- markov random field
- computer vision
- input image
- high level vision
- image analysis
- single image
- multiscale
- image retrieval
- energy function
- region of interest
- image content
- image collections
- pixel values
- image representation
- image features
- low level
- high resolution
- object recognition
- fuzzy logic
- image regions
- image set
- image segmentation
- natural images
- keypoints
- test images
- training set
- computational complexity
- image pixels
- similarity measure