Anna Belardinelli, Chao Wang, Michael Gienger, "Explainable Human Robot Interaction for Imitation Learning in Augmented Reality", Horizons of an Extended Robotics Reality (XR2) (workshop @IROS2022), 2022.

Imitation learning could enable non-expert users to teach new skills to robots in an interactive and intuitive way. Still, it is often diffi cult to grasp what the robot knows or to assess if a correct representation of the task is being formed. Here, we introduce an Explainable AI (XAI) design for human-robot interaction during learning by demonstration of simple kitchen tasks in Augmented Reality. AR-XAI cues are used to visualize the perceptual beliefs of the robot in an interactive way. These are indeed overlaid directly on the shared workspace, as perceived by the teacher, in so letting the teacher access the robot’s situation understanding without the need for an explicit inquiry. We conducted a user study to assess the benefi ts of such feedback modality, as compared to a more human-like feedback modality such as gaze and speech, and to a combination of both behavioral and virtual cues. AR-XAI cues are well received, still they enhance the robot perceived competence only in combination with more social cues.