Learning Visual Relation Priors for Image-Text Matching and Image Captioning with Neural Scene Graph Generators.
Kuang-Huei LeeHamid PalangiXi ChenHoudong HuJianfeng GaoPublished in: CoRR (2019)
Keyphrases
- input image
- single image
- image data
- image features
- multiscale
- image content
- template matching
- image classification
- keypoints
- image collections
- low level
- visual appearance
- feature points
- web images
- image regions
- image representation
- scene images
- scene matching
- image retrieval
- image matching
- outdoor scenes
- multiple objects
- auto annotation
- high resolution
- image segmentation
- multiple images
- visual data
- complex scenes
- spatial information
- visual features
- point features
- spatial relations
- graph matching
- bayesian framework
- computer vision
- object detection
- image derivatives