Modeling Explicit Concerning States for Reinforcement Learning in Visual Dialogue.
Zipeng XuFandong MengXiaojie WangDuo ZhengChenxu LvJie ZhouPublished in: BMVC (2021)
Keyphrases
- perceptual aliasing
- reinforcement learning
- function approximation
- visual features
- visual perception
- low level
- state space
- visual information
- multi agent
- learning algorithm
- dynamic programming
- supervised learning
- machine learning
- multi agent systems
- dialogue system
- visual representation
- mixed initiative
- practical reasoning
- dialogue management