To effectively participate in a group discussion, it's important to be able to identify who is present and direct your attention accordingly. For most people, this is not hard, but designing robots able to do the same thing is quite challenging. While robots are equipped with sensors for detecting the number of people in a group, they are听not always accurate and, to date, there has been little research into how robots can confirm their assumptions and correct any errors they may have made.
However, last year, two researchers听in ATLAS Institute's听IRON Lab* developed a solution to this problem that is described in听a paper听published in the March Proceedings of the听International Conference on Human-Robot Interaction 听(HRI '22). The authors,听Hooman Hedayati听听(PhD computer science '20) and Daniel Szafir, assistant professor of computer science at UNC Chapel Hill听and the former director of the ATLAS IRON Lab, proposed a method to overcome situations when conversational group (F-formation) detection algorithms fail.
By studying different conversational group听data sets, the researchers observed that relative to the size of a conversation group, people tend to stand in听predictable locations relative to each other. Hedayati and Szafir听then developed a system for identifying high probability regions where people are likely to stand in a group relative to a single participant.听Using that system, the robot can reason when another person in the听conversation hasn't been detected and correct their error.听
The first model estimates the true size of a conversational group, where only some participants were detected. The听second model predicts the locations where any undetected participants are likely to be standing. Together, these models may improve detection algorithms and a robot's ability to detect members of a group and participate more seamlessly in a conversation.
*Following Szafir's departure last summer, the ATLAS听IRON Lab was closed.
听
Publication
Hooman Hedayati and Daniel Szafir. 2022. . In Proceedings of the 2022 ACM/IEEE International Conference on Human-Robot Interaction (HRI '22). IEEE Press, 402鈥�411 (March 7-10, 2022鈥攙irtual, originally Hokkaido, Japan).
