Make A Long Image Short: Adaptive Token Length for Vision Transformers.
Qiqi ZhouYichen ZhuPublished in: CoRR (2023)
Keyphrases
- single image
- image data
- template matching
- image pixels
- multiscale
- input image
- image content
- keypoints
- image analysis
- visual perception
- low level image processing
- image classification
- image representation
- grey level
- image synthesis
- spatial information
- image features
- low level
- real time
- image matching
- image segmentation
- pixel values
- segmentation method
- image collections
- lighting conditions
- hough transform
- image regions
- vision system
- image retrieval
- similarity measure
- computer vision