RelViT: Concept-guided Vision Transformer for Visual Relational Reasoning.
Xiaojian MaWeili NieZhiding YuHuaizu JiangChaowei XiaoYuke ZhuSong-Chun ZhuAnima AnandkumarPublished in: CoRR (2022)
Keyphrases
- visual perception
- visual processing
- human vision
- computer vision
- visual field
- low level
- visual information
- real time
- power system
- model based reasoning
- qualitative reasoning
- probabilistic reasoning
- visual cues
- visual data
- concept learning
- relational data
- vision system
- fuzzy logic
- database systems
- databases
- visual features
- knowledge representation
- data model
- relational databases
- visual concepts
- knowledge base
- artificial intelligence
- reasoning tasks
- reasoning process
- machine learning
- neural network