Customizable Text-to-Image Modeling by Contrastive Learning on Adjustable Word-Visual Pairs.
Jun-Li LuYoichi OchiaiPublished in: HCI (35) (2022)
Keyphrases
- auto annotation
- learning process
- image analysis
- learning algorithm
- printed documents
- image data
- image features
- input image
- single image
- information retrieval
- automatic image annotation
- image classification
- image representation
- image retrieval
- image regions
- word pairs
- visual appearance
- visual perception
- web images
- edge detection
- low level
- multiscale
- visual cues
- visual data
- test images
- document analysis
- relevance feedback
- visual similarity
- information extraction