Reinforcement-learning based dialogue system for human-robot interactions with socially-inspired rewards.
Emmanuel FerreiraFabrice LefèvrePublished in: Comput. Speech Lang. (2015)
Keyphrases
- human robot
- dialogue system
- reinforcement learning
- action selection
- human computer
- multi agent
- dialogue management
- natural language generation
- markov decision processes
- natural language
- spoken dialogue systems
- state space
- spoken language
- tutorial dialogue
- learning algorithm
- model free
- human users
- human robot interaction
- language understanding
- robot behavior
- reinforcement learning algorithms
- optimal policy
- reward function
- partially observable
- dynamic programming
- real time
- computer vision
- participatory simulation
- partially observable markov decision processes
- multiple agents
- humanoid robot
- collaborative filtering
- domain knowledge
- cooperative
- image sequences