Development of a Multi-sensor Emotional Response System for Social Robots

Loneliness is a significant and prevalent public health issue. To combat this, many social intervention agents have been developed the majority of these systems limit themselves to one emotional categorization system and as a result, lack a complex emotional understanding of the user. This paper has created a novel system for use in companion robots. By using sensor fusion of touch-sensitive fur and computer vision technology it determines the emotional state of the user. Combining these sensor inputs help deepen the estimate of the emotional state of the user and allows for the system to predict more complex emotions. This paper is aimed to help create a companion robot that responds to the user in a sympathetic way that will best create a positive response in the user. As an additional result, the achieved touch classification accuracy was found high enough to show promise for this innovative approach to touch sensing in the future.

Automated summary

The paper introduces a novel companion robot system that uses sensor fusion of touch-sensitive fur and computer vision technology to determine and predict user emotions, aiming to respond sympathetically and elicit a positive response from the user. Additionally, the achieved touch classification accuracy shows promise for future applications of this innovative approach to touch sensing.