Inner Speech and Damasio’s Theory for Modeling Robot’s Emotions

摘要

In Affective Robotics, there is a growing emphasis on endowing robots with the ability to experience emotions, not just detect and recognise human emotions. Robots can appropriately respond to emotionally relevant events, simulating emotional behaviour, thus improving social interaction. This work analyses the robot's emotional experiences by inner speech. Recent investigations demonstrated that the robot's inner speech improves people's trust and brings the robot closer to human cognition. Through inner speech, the robot overtly talks to itself and reasons about the context and inner processes. Based on Damasio's theory, emotions emerge from implementing the dynamic interplay between bodily sensations and emotional cognitive processes. Integrating such a theory with self-talking capability and deploying the resulting model on a real robot makes such a robot able to experience emotions. Experiments demonstrate that people interacting with robots equipped with the proposed model recognise the robot's emotional experiences, supporting a more empathetic and emotionally connected relationship.