MKVSE: Multimodal Knowledge Enhanced Visual-semantic Embedding for Image-text Retrieval.
Duoduo FengXiangteng HeYuxin PengPublished in: ACM Trans. Multim. Comput. Commun. Appl. (2023)
Keyphrases
- text retrieval
- image retrieval
- object retrieval
- image features
- image content
- image representation
- information retrieval
- semantic information
- document collections
- retrieval systems
- low level
- vector space
- visual concepts
- visual features
- image classification
- document retrieval
- medical image retrieval
- visual similarity
- query expansion
- visually similar
- semantic similarity
- image collections
- text queries
- image descriptors
- inverted file
- bag of visual words
- cross language
- retrieval model
- visual information
- image regions
- latent semantic indexing
- web images
- information retrieval systems
- handwritten documents
- natural language
- automatic query expansion
- similarity measure