PERSEO at IROS 2022, International Conference on Robots and Intelligent Systems in Kyoto, Japan.
Last week PERSEO, with three early-stage researchers and their supervisors, participated in the IEEE/RSJ International Conference on Robots and Intelligent Systems in Kyoto, Japan.
Georgios Angelopoulos presented the paper “You Are in My Way: Non-Verbal Social Cues for Legible Robot Navigation Behaviors”. This work focuses on integrating legible non-verbal behaviours into the robot’s social navigation to make nearby humans aware of its intended trajectory. Results from a within-subjects study involving 33 participants show that deictic gestures as navigational cues for humanoid robots result in fewer navigation conflicts than the use of a simulated gaze. Additionally, an increase in the perceived anthropomorphism is found when the robot uses the deictic gesture as a cue. These findings show the importance of social behaviours for people avoidance and suggest a paradigm of such behaviours in future humanoid robotic applications.
Esteve Valls Mascaro presented the paper “Robust Human Motion Forecasting Using Transformer-Based Model”. In this work, they propose a new model based on Transformer that simultaneously deals with the real-time 3D human motion forecasting in the short and long term. Their 2-Channel Transformer (2CH-TR) is able to efficiently exploit the spatio-temporal information of a shortly observed sequence (400ms) and generates a competitive accuracy against the current state-of-the-art. 2CH-TR stands out for the efficient performance of the Transformer, being lighter and faster than its competitors. In addition, the model is tested in conditions where the human motion is severely occluded, demonstrating its robustness in reconstructing and predicting 3D human motion in a highly noisy environment. Their experiment results show that the proposed 2CH-TR outperforms the ST-Transformer, which is another state-of-the-art model based on the Transformer, in terms of reconstruction and prediction under the same conditions of input prefix.
The model reduces in 8.89% the mean squared error of ST-Transformer in short-term prediction, and 2.57% in long-term prediction in the Human3.6M dataset with 400ms input prefix.
Lorenzo Ferrini co-organized the “ROS4HRI Tutorial: From Zero to Multi-Modal Interactive Dialogue for Robots” tutorial. During this one-day tutorial, they presented the ROS4HRI standard and instructed the attendees on how to program a small simulated interactive system able to track and react to faces, skeletons and chats.
