Multi-Scale Vision Longformer: A New Vision Transformer for High-Resolution Image Encoding.
Pengchuan ZhangXiyang DaiJianwei YangBin XiaoLu YuanLei ZhangJianfeng GaoPublished in: CoRR (2021)
Keyphrases
- high resolution
- multiscale
- image processing
- low level image processing
- image representation
- computer vision
- low resolution
- edge detection
- visual perception
- input image
- image classification
- image segmentation
- lower resolution
- low resolution images
- image features
- image analysis
- high resolution images
- multiple scales
- vision system
- super resolution
- image content
- segmentation method
- single image
- image data
- real time
- high frequency
- test images
- satellite images
- wavelet decomposition
- feature points
- spatial color
- high quality
- image generation
- super resolution reconstruction
- visual field
- remote sensing
- image synthesis
- natural images
- segmentation algorithm
- scale space
- higher resolution
- coarse to fine
- field of view
- image regions
- keypoints
- wavelet coefficients