Policy Optimization to Learn Adaptive Motion Primitives in Path Planning With Dynamic Obstacles | IEEE Journals & Magazine | IEEE Xplore