Erwin Coumans
YOU?
Author Swipe
View article: MaskedManipulator: Versatile Whole-Body Manipulation
MaskedManipulator: Versatile Whole-Body Manipulation Open
We tackle the challenges of synthesizing versatile, physically simulated human motions for full-body object manipulation. Unlike prior methods that are focused on detailed motion tracking, trajectory following, or teleoperation, our framew…
View article: Achieving Human Level Competitive Robot Table Tennis
Achieving Human Level Competitive Robot Table Tennis Open
Achieving human-level speed and performance on real world tasks is a north star for the robotics research community. This work takes a step towards that goal and presents the first learned robot agent that reaches amateur human-level perfo…
View article: Robotic Table Tennis: A Case Study into a High Speed Learning System
Robotic Table Tennis: A Case Study into a High Speed Learning System Open
We present a deep-dive into a real-world robotic learning system that, in previous work, was shown to be capable of hundreds of table tennis rallies with a human and has the ability to precisely return the ball to desired targets. This sys…
View article: Robotic Table Tennis: A Case Study into a High Speed Learning System
Robotic Table Tennis: A Case Study into a High Speed Learning System Open
We present a deep-dive into a real-world robotic learning system that, in previous work, was shown to be capable of hundreds of table tennis rallies with a human and has the ability to precisely return the ball to desired targets. This sys…
View article: Barkour: Benchmarking Animal-level Agility with Quadruped Robots
Barkour: Benchmarking Animal-level Agility with Quadruped Robots Open
Animals have evolved various agile locomotion strategies, such as sprinting, leaping, and jumping. There is a growing interest in developing legged robots that move like their biological counterparts and show various agile skills to naviga…
View article: Differentiable Dynamics for Articulated 3d Human Motion Reconstruction
Differentiable Dynamics for Articulated 3d Human Motion Reconstruction Open
We introduce DiffPhy, a differentiable physics-based model for articulated 3d human motion reconstruction from video. Applications of physics-based reasoning in human motion analysis have so far been limited, both by the complexity of cons…
View article: Inferring Articulated Rigid Body Dynamics from RGBD Video
Inferring Articulated Rigid Body Dynamics from RGBD Video Open
Being able to reproduce physical phenomena ranging from light interaction to contact mechanics, simulators are becoming increasingly useful in more and more application domains where real-world interaction or labeled data are difficult to …
View article: Braxlines: Fast and Interactive Toolkit for RL-driven Behavior Engineering beyond Reward Maximization
Braxlines: Fast and Interactive Toolkit for RL-driven Behavior Engineering beyond Reward Maximization Open
The goal of continuous control is to synthesize desired behaviors. In reinforcement learning (RL)-driven approaches, this is often accomplished through careful task reward engineering for efficient exploration and running an off-the-shelf …
View article: Multi-Task Learning with Sequence-Conditioned Transporter Networks
Multi-Task Learning with Sequence-Conditioned Transporter Networks Open
Enabling robots to solve multiple manipulation tasks has a wide range of industrial applications. While learning-based approaches enjoy flexibility and generalizability, scaling these approaches to solve such compositional tasks remains a …
View article: NeuralSim: Augmenting Differentiable Simulators with Neural Networks
NeuralSim: Augmenting Differentiable Simulators with Neural Networks Open
Differentiable simulators provide an avenue for closing the sim-to-real gap by enabling the use of efficient, gradient-based optimization algorithms to find the simulation parameters that best fit the observed sensor readings. Nonetheless,…
View article: Fast and Efficient Locomotion via Learned Gait Transitions
Fast and Efficient Locomotion via Learned Gait Transitions Open
We focus on the problem of developing energy efficient controllers for quadrupedal robots. Animals can actively switch gaits at different speeds to lower their energy consumption. In this paper, we devise a hierarchical learning framework,…
View article: Learning to Rearrange Deformable Cables, Fabrics, and Bags with Goal-Conditioned Transporter Networks
Learning to Rearrange Deformable Cables, Fabrics, and Bags with Goal-Conditioned Transporter Networks Open
Rearranging and manipulating deformable objects such as cables, fabrics, and bags is a long-standing challenge in robotic manipulation. The complex dynamics and high-dimensional configuration spaces of deformables, compared to rigid object…
View article: Augmenting Differentiable Simulators with Neural Networks to Close the Sim2Real Gap
Augmenting Differentiable Simulators with Neural Networks to Close the Sim2Real Gap Open
We present a differentiable simulation architecture for articulated rigid-body dynamics that enables the augmentation of analytical models with neural networks at any point of the computation. Through gradient-based optimization, identific…
View article: Learning Agile Robotic Locomotion Skills by Imitating Animals
Learning Agile Robotic Locomotion Skills by Imitating Animals Open
Reproducing the diverse and agile locomotion skills of animals has been a longstanding challenge in robotics. While manually-designed controllers have been able to emulate many complex behaviors, building such controllers involves a time-c…
View article: Learning Agile Robotic Locomotion Skills by Imitating Animals
Learning Agile Robotic Locomotion Skills by Imitating Animals Open
Reproducing the diverse and agile locomotion skills of animals has been a longstanding challenge in robotics. While manually-designed controllers have been able to emulate many complex behaviors, building such controllers involves a time-c…
View article: Policies Modulating Trajectory Generators
Policies Modulating Trajectory Generators Open
We propose an architecture for learning complex controllable behaviors by having simple Policies Modulate Trajectory Generators (PMTG), a powerful combination that can provide both memory and prior knowledge to the controller. The result i…
View article: Learning Fast Adaptation with Meta Strategy Optimization
Learning Fast Adaptation with Meta Strategy Optimization Open
The ability to walk in new scenarios is a key milestone on the path toward real-world applications of legged robots. In this work, we introduce Meta Strategy Optimization, a meta-learning algorithm for training policies with latent variabl…
View article: Sim-to-Real: Learning Agile Locomotion For Quadruped Robots
Sim-to-Real: Learning Agile Locomotion For Quadruped Robots Open
Designing agile locomotion for quadruped robots often requires extensive expertise and tedious manual tuning.In this paper, we present a system to automate this process by leveraging deep reinforcement learning techniques.Our system can le…
View article: Optimizing Simulations with Noise-Tolerant Structured Exploration
Optimizing Simulations with Noise-Tolerant Structured Exploration Open
We propose a simple drop-in noise-tolerant replacement for the standard finite difference procedure used ubiquitously in blackbox optimization. In our approach, parameter perturbation directions are defined by a family of structured orthog…
View article: Sim-to-Real: Learning Agile Locomotion For Quadruped Robots
Sim-to-Real: Learning Agile Locomotion For Quadruped Robots Open
Designing agile locomotion for quadruped robots often requires extensive expertise and tedious manual tuning. In this paper, we present a system to automate this process by leveraging deep reinforcement learning techniques. Our system can …