VICTR: Visual Information Captured Text Representation for Text-to-Image Multimodal Tasks.
Soyeon Caren HanSiqu LongSiwen LuoKunze WangJosiah PoonPublished in: CoRR (2020)
Keyphrases
- visual information
- textual information
- text representation
- low level
- visual data
- visual features
- image features
- image classification
- text documents
- image retrieval
- visual content
- image representation
- text retrieval
- multiscale
- text mining
- image content
- eye movements
- concept learning
- machine learning
- statistical natural language processing
- document clustering
- domain knowledge
- similarity measure
- information retrieval
- knn
- relational databases