VISTA: Visualized Text Embedding For Universal Multi-Modal Retrieval.
Junjie ZhouZheng LiuShitao XiaoBo ZhaoYongping XiongPublished in: ACL (1) (2024)
Keyphrases
- multi modal
- video search
- cross modal
- information retrieval
- text retrieval
- multiple modalities
- high dimensional
- audio visual
- text mining
- multi modality
- keywords
- multimedia documents
- relevance feedback
- image retrieval
- text documents
- vector space
- semantic concepts
- multimedia retrieval
- auto annotation
- retrieval systems
- magnetic resonance images
- content based retrieval
- image annotation
- information retrieval systems
- co occurrence
- similarity measure
- face recognition
- feature selection
- search engine