-
End-to-End Motion Planning of Quadrotors Using Deep Reinforcement Learning
The proposed method uses raw depth images obtained from a front-facing camera and directly generates local motion plans in the form of smooth motion primitives that move a...