Weakly supervised grounded image captioning with semantic matching.
Sen DuHong ZhuGuangfeng LinYuanyuan LiuDong WangJing ShiZhong WuPublished in: Appl. Intell. (2024)
Keyphrases
- weakly supervised
- input image
- image features
- semantic matching
- image retrieval
- multiscale
- single image
- image classification
- image representation
- image segmentation
- image content
- region of interest
- data sets
- viewpoint
- energy function
- segmentation method
- test images
- multiple images
- relation extraction
- training set
- ontology alignment