An Emotion-Based Interaction Strategy to Improve Human-Robot Interaction

Abstract

Emotion and empathy are key subjects on human-robot interaction, especially regarding social robots. Several studies have investigated emotional reactions of humans toward robots, while others deal with development of actual systems to analyze how affective feedback may influence this kind of interaction. This paper presents an emotion-aware interaction strategy applied to an embodied virtual agent, implemented as an Android application. The system assigns two distinct paradigms to the virtual character, according to the user's emotion, inferred through facial expressions analysis. Within subject user experiments have been performed, in order to evaluate if the proposed strategy improves empathy and pleasantness.