ERM: Energy-Based Refined-Attention Mechanism for Video Question Answering.
Fuwei ZhangRuomei WangFan ZhouYuanmao LuoPublished in: IEEE Trans. Circuits Syst. Video Technol. (2023)
Keyphrases
- question answering
- attention mechanism
- natural language processing
- video sequences
- question classification
- video data
- video streams
- information retrieval
- multimedia
- information extraction
- visual attention
- cross language
- qa clef
- syntactic information
- natural language
- video content
- passage retrieval
- natural language questions
- video frames
- key frames
- video retrieval
- visual attention model
- answer validation
- question answering systems
- semantic roles
- video summarization
- qa systems
- answer extraction
- saliency map
- wordnet