Self-supervised multi-scale semantic consistency regularization for unsupervised image-to-image translation.
Heng ZhangYi-Jun YangWei ZengPublished in: Comput. Vis. Image Underst. (2024)
Keyphrases
- multiscale
- image data
- input image
- image content
- image representation
- blind deconvolution
- image segmentation
- image classification
- image analysis
- template matching
- single image
- low level
- image pixels
- image collections
- high resolution
- image retrieval
- hough transform
- spatial information
- image features
- keypoints
- segmentation method
- image regions
- region of interest
- image structure
- image matching
- test images
- segmentation algorithm
- edge detection
- supervised learning
- object recognition