Empathetic Robot With Transformer-Based Dialogue Agent

Abstract

Natural human-robot interaction (HRI) attracts considerable interest in letting robots understand users’ emotional states. This paper demonstrates a method to introduce the affection model to a robotic system’s conversational agent to provide natural and empathetic HRI. We use a large-scale pre-trained language model and fine-tune it on a dialogue dataset with empathetic characteristics. Based on recent progress in deep learning and natural language processing, we extend the current methods and enable our robotic agent to perform advanced sentiment analysis using our trained affection model. This dialogue agent will allow the robot to provide natural responses along with emotion classification with the estimations of arousal and valence levels. We evaluate our model using different metrics in comparison with recent studies and showing its emotion detection capability.