Towards Context Aware Emotion Recognition in HRI for Social Robots
- Publication Type:
- Conference Proceeding
- Citation:
- Australasian Conference on Robotics and Automation, ACRA, 2023
- Issue Date:
- 2023-01-01
Open Access
Copyright Clearance Process
- Recently Added
- In Progress
- Open Access
This item is open access.
Social robots are becoming more prevalent in our daily environments but continue to struggle communicating in human-robot interactions, often misunderstanding people and thus making the interaction uncomfortable. Many attempts have been made to improve their understanding of people and their emotions but they still lack the socio-emotional intelligence humans often use in human-human interactions. A new approach previously explored in computer science is using context emotion recognition to interpret a scene for clues to a person’s emotional state. In this paper, we state that context emotion recognition will benefit the fields of human-robot interaction and social robotics. Further, we extend upon the EMOTIC model successfully adding a graphical representation of the emotion probabilities over time to the model output and with the addition of a facial feature extractor module that obtains an encouraging improvement over the original model. The algorithm was validated through data coming from two robotic platforms, namely PR2 and Pepper. The results show a promising way towards context aware emotion recognition in human-robot interactions with social robots, with 88% accuracy when comparing with 66% accuracy of the base model.
Please use this identifier to cite or link to this item: