GraDual: Graph-based Dual-modal Representation for Image-Text Matching.
Siqu LongSoyeon Caren HanXiaojun WanJosiah PoonPublished in: WACV (2022)
Keyphrases
- image representation
- template matching
- image data
- single image
- image analysis
- feature points
- image matching
- input image
- keypoints
- matching scheme
- matching process
- web images
- multiscale
- image features
- feature matching
- scene matching
- image classification
- test images
- image set
- image content
- similarity measure
- edge detection
- feature descriptors
- image retrieval
- feature representation
- spatial pyramid matching
- false matches
- object matching
- textual descriptions
- low level
- normalized correlation
- spatial arrangement
- spatial relations
- region of interest
- image segmentation
- spatial information
- bag of words
- image pixels
- high resolution
- sift descriptors
- semi supervised
- shape matching
- image collections
- pattern matching