Pre-Training Multi-Modal Dense Retrievers for Outside-Knowledge Visual Question Answering.
Alireza SalemiMahta RafieeHamed ZamaniPublished in: ICTIR (2023)
Keyphrases
- multi modal
- question answering
- cross modal
- video search
- multi modality
- natural language processing
- question classification
- knowledge base
- knowledge representation
- high dimensional
- syntactic information
- natural language questions
- qa clef
- cross language
- audio visual
- natural language
- information retrieval
- question answering systems
- visual information
- passage retrieval
- information extraction
- semantic roles
- expert systems
- training set
- keywords
- candidate answers
- single modality
- answer validation
- image annotation
- vector space
- semantic information
- domain knowledge