ReFormer: The Relational Transformer for Image Captioning.
Xuewen YangYingru LiuXin WangPublished in: CoRR (2021)
Keyphrases
- image features
- image data
- input image
- single image
- image classification
- image analysis
- image segmentation
- image regions
- image representation
- multiscale
- image content
- spatial information
- low level
- image retrieval
- region of interest
- grey level
- image pixels
- image collections
- segmentation method
- computer vision
- feature points
- image structure
- pixel values
- fuzzy logic
- image matching
- keypoints
- test images
- relational databases