Neural Dynamic Policies for End-to-End Sensorimotor Learning
NIPS 2020, 2020.
We propose a novel re-parameterization of action spaces via Neural Dynamic Policies, a set of policies which impose the structure of a dynamical system on action spaces
The current dominant paradigm in sensorimotor control, whether imitation or reinforcement learning, is to train policies directly in raw action spaces such as torque, joint angle, or end-effector position. This forces the agent to make decisions individually at each timestep in training, and hence, limits the scalability to continuous, ...More
PPT (Upload PPT)