Improved Multi-Agent Deep Deterministic Policy Gradient for Path Planning-Based Crowd Simulation | IEEE Journals & Magazine | IEEE Xplore