The Robot-Pedestrian Influence Dataset for Learning Distinct Social Navigation Forces

Existing research lacks comprehensive datasets that capture the full range of pedestrian behaviors, e.g., including avoidance, neutrality, and attraction in the presence of robots. In this paper, we present a novel dataset capturing pedestrian behavior in the presence of robots under varying conditions, enabling better prediction of responses like avoidance or attraction. Leveraging this, we introduce the Neural Social Robot Force Model (NSRFM), which outperforms baselines in predicting real-world pedestrian trajectories and supports the

development of socially-aware robot navigation.

Citation information

Agrawal, Subham; Ostermann-Myrau, Nico; Dengler, Nils; Bennewitz, Maren: The Robot-Pedestrian Influence Dataset for Learning Distinct Social Navigation Forces, Proceedings of 7th Workshop on Long-term Human Motion Prediction, 2025, May, https://motionpredictionicra2025.github.io/assets/papers/Agrawal2025.pdf, Agrawal.etal.2025b,