Natural Language Rationales with Full-Stack Visual Reasoning: From Pixels to Semantic Frames to Commonsense Graphs.
Ana MarasovicChandra BhagavatulaJae Sung ParkRonan Le BrasNoah A. SmithYejin ChoiPublished in: EMNLP (Findings) (2020)
Keyphrases
- natural language
- commonsense knowledge
- knowledge representation
- language understanding
- natural language understanding
- semantic interpretation
- semantic analysis
- knowledge base
- semantic representation
- natural language processing
- commonsense reasoning
- linguistic analysis
- conceptual graphs
- natural language interface
- machine learning
- graph matching
- language processing
- visual features
- logical formulas
- automated reasoning
- dialogue system
- question answering
- semantic network
- semantic information
- visual information
- graph model
- legal cases
- natural language sentences
- consecutive frames
- semantic space
- directed graph
- semantically relevant
- video frames
- moving objects
- information extraction
- visual concepts
- frame rate
- input image
- low level
- key frames
- low level features
- semantic content
- syntactic structures
- semantic similarity
- adjacency graph
- graph mining