As robots interact with people, people form opinions of the robots' capabilities. In this project, we are interested in interpreting people's movements and actions as passive communication about what they think about the robots they are interacting with, so that robots can understand what people think of it. The information can be used to help with building trust and improving teamwork. 



  • Connor Brooks and Daniel Szafir. 2019. Building Second-Order Mental Models for Human-Robot Interaction. In Proceedings of the Association for the Advancement of Artificial Intelligence Fall Symposium Series (AI-HRI '19), ArXiv preprint arXiv:1909.06508.