Shared Space

Shared space research focuses on modeling and optimizing interactions within environments where multiple agents, such as humans, autonomous vehicles, and robots, coexist. Current efforts concentrate on improving prediction accuracy of agent behavior (e.g., pedestrian trajectories) using deep reinforcement learning and novel interaction models like polar collision grids, often incorporating uncertainty quantification to enhance safety and efficiency. This work has significant implications for the development of safe and efficient autonomous systems, particularly in crowded urban settings, and for improving human-computer interaction in shared digital spaces like multilingual text processing.