Neural Dynamic Policies for End-to-End Sensorimotor Learning

December 6, 2020


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 decision at each point in training, and hence, limit the scalability to continuous, high-dimensional, and long-horizon tasks. In contrast, research in classical robotics has, for a long time, exploited dynamical systems as a policy representation to learn robot behaviors via demonstrations. These techniques, however, lack the flexibility and generalizability provided by deep learning or deep reinforcement learning and have remained under-explored in such settings. In this work, we begin to close this gap and embed dynamics structure into deep neural network-based policies by reparameterizing action spaces with differential equations. We propose Neural Dynamic Policies (NDPs) that make predictions in trajectory distribution space as opposed to prior policy learning methods where action represents the raw control space. The embedded structure allow us to perform end-to-end policy learning under both reinforcement and imitation learning setups. We show that NDPs achieve better or comparable performance to state-of-the-art approaches on many robotic control tasks using both reward-based training and demonstrations. Project video and code are available at: neural-dynamic-policies/

Download the Paper


Written by

Shikhar Bahl

Mustafa Mukadam

Abhinav Gupta

Deepak Pathak

Research Topics


Reinforcement Learning

Related Publications

June 15, 2019


FBNet: Hardware-Aware Efficient ConvNet Design via Differentiable Neural Architecture Search | Facebook AI Research

Designing accurate and efficient ConvNets for mobile devices is challenging because the design space is combinatorially large. Due to this, previous neural architecture search (NAS) methods are computationally expensive. ConvNet architecture…

Bichen Wu, Xiaoliang Dai, Peizhao Zhang, Yanghan Wang, Fei Sun, Yiming Wu, Yuandong Tian, Peter Vajda, Yangqing Jia, Kurt Keutzer

June 15, 2019

April 28, 2019


Inverse Path Tracing for Joint Material and Lighting Estimation | Facebook AI Research

Modern computer vision algorithms have brought significant advancement to 3D geometry reconstruction. However, illumination and material reconstruction remain less studied, with current approaches assuming very simplified models for materials…

Dejan Azinović, Tzu-Mao Li, Anton Kaplanyan, Matthias Nießner

April 28, 2019

June 14, 2019


Thinking Outside the Pool: Active Training Image Creation for Relative Attributes | Facebook AI Research

Current wisdom suggests more labeled image data is always better, and obtaining labels is the bottleneck. Yet curating a pool of sufficiently diverse and informative images is itself a challenge. In particular, training image curation is…

Aron Yu, Kristen Grauman

June 14, 2019

September 09, 2018


DDRNet: Depth Map Denoising and Refinement for Consumer Depth Cameras Using Cascaded CNNs | Facebook AI Research

Consumer depth sensors are more and more popular and come to our daily lives marked by its recent integration in the latest iPhone X. However, they still suffer from heavy noises which dramatically limit their applications. Although plenty of…

Shi Yan, Chenglei Wu, Lizhen Wang, Feng Xu, Liang An, Kaiwen Guo, Yebin Liu

September 09, 2018

Help Us Pioneer The Future of AI

We share our open source frameworks, tools, libraries, and models for everything from research exploration to large-scale production deployment.