An Interactive Humanoid Robot Exhibiting Flexible Sub-Dialogues

We demonstrate a conversational humanoid robot that allows users to follow their own dialogue structures. Our system uses a hierarchy of reinforcement learning dialogue agents, which support transitions across sub-dialogues in order to relax the strictness of hierarchical control and therefore support flexible interactions. We demonstrate our system with the Nao robot playing two versions of a Quiz game. Whilst language input and dialogue control is autonomous or wizarded, language output is provided by the robot combining verbal and non-verbal contributions. The novel features in our system are (a) the flexibility given to users to navigate flexibly in the interaction; and (b) a framework for investigating adaptive and flexible dialogues.