SwapMix: Diagnosing and Regularizing the Over-Reliance on Visual Context in Visual Question Answering.
Vipul GuptaZhuowan LiAdam KortylewskiChenyu ZhangYingwei LiAlan L. YuillePublished in: CVPR (2022)
Keyphrases
- question answering
- visual context
- semantic context
- visual scene
- temporal context
- natural language
- information retrieval
- object detection
- scene interpretation
- information extraction
- visual information
- natural language processing
- co occurrence
- visual words
- high level
- question answering systems
- video sequences
- low level
- keywords
- artificial intelligence