DynGraph: Visual Question Answering via Dynamic Scene Graphs.
Monica HauriletZiad Al-HalahRainer StiefelhagenPublished in: GCPR (2019)
Keyphrases
- question answering
- dynamic scenes
- space time
- multi view
- video sequences
- cross language
- passage retrieval
- information retrieval
- background subtraction
- question classification
- question answering systems
- natural language processing
- natural language questions
- natural language
- visual features
- qa clef
- information extraction
- moving objects
- motion segmentation
- syntactic information
- frame rate
- graph structure
- data mining
- computer vision
- combining information from multiple
- semantic roles
- visual information
- document collections
- knowledge representation
- knowledge base