MSCAN: Multimodal Self-and-Collaborative Attention Network for image aesthetic prediction tasks.
Xiaodan ZhangXinbo GaoLihuo HeWen LuPublished in: Neurocomputing (2021)
Keyphrases
- image data
- single image
- input image
- multiscale
- template matching
- segmentation method
- image features
- image representation
- image pixels
- high resolution
- image analysis
- image segmentation
- image classification
- multi modal
- image collections
- image content
- low level
- segmentation algorithm
- image regions
- image retrieval
- region of interest
- color images
- lighting conditions
- image matching
- peer to peer
- test images
- keypoints
- computer vision