SegEQA: Video Segmentation Based Visual Attention for Embodied Question Answering.
Haonan LuoGuosheng LinZichuan LiuFayao LiuZhenmin TangYazhou YaoPublished in: ICCV (2019)
Keyphrases
- question answering
- video segmentation
- visual attention
- eye tracking
- video sequences
- eye movements
- video frames
- saliency map
- visual search
- vision system
- information extraction
- natural language processing
- higher level
- information retrieval
- natural language
- video analysis
- segmentation method
- salient regions
- qa clef
- visual motion
- image data
- low level
- image processing
- real time
- human computer interaction
- visual saliency
- question answering systems
- natural language questions