Learning Cross-Domain Correspondence for Control with Dynamics Cycle-Consistency

• URL: http://arxiv.org/abs/2012.09811v1
• Date: Thu, 17 Dec 2020 18:22:25 GMT
• Title: Learning Cross-Domain Correspondence for Control with Dynamics Cycle-Consistency
• Authors: Qiang Zhang, Tete Xiao, Alexei A. Efros, Lerrel Pinto, Xiaolong Wang
• Abstract summary: We learn to align dynamic robot behavior across two domains using a cycle-consistency constraint. Our framework is able to align uncalibrated monocular video of a real robot arm to dynamic state-action trajectories of a simulated arm without paired data.
• Score: 60.39133304370604
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: At the heart of many robotics problems is the challenge of learning correspondences across domains. For instance, imitation learning requires obtaining correspondence between humans and robots; sim-to-real requires correspondence between physics simulators and the real world; transfer learning requires correspondences between different robotics environments. This paper aims to learn correspondence across domains differing in representation (vision vs. internal state), physics parameters (mass and friction), and morphology (number of limbs). Importantly, correspondences are learned using unpaired and randomly collected data from the two domains. We propose \textit{dynamics cycles} that align dynamic robot behavior across two domains using a cycle-consistency constraint. Once this correspondence is found, we can directly transfer the policy trained on one domain to the other, without needing any additional fine-tuning on the second domain. We perform experiments across a variety of problem domains, both in simulation and on real robot. Our framework is able to align uncalibrated monocular video of a real robot arm to dynamic state-action trajectories of a simulated arm without paired data. Video demonstrations of our results are available at: https://sjtuzq.github.io/cycle_dynamics.html .

Related papers

• Dynamics as Prompts: In-Context Learning for Sim-to-Real System Identifications [23.94013806312391]
  We propose a novel approach that dynamically adjusts simulation environment parameters online using in-context learning. We validate our approach across two tasks: object scooping and table air hockey. Our approach delivers efficient and smooth system identification, advancing the deployment of robots in dynamic real-world scenarios.
  arXiv  Detail & Related papers  (2024-10-27T07:13:38Z)
• Flow as the Cross-Domain Manipulation Interface [73.15952395641136]
  Im2Flow2Act enables robots to acquire real-world manipulation skills without the need of real-world robot training data. Im2Flow2Act comprises two components: a flow generation network and a flow-conditioned policy. We demonstrate Im2Flow2Act's capabilities in a variety of real-world tasks, including the manipulation of rigid, articulated, and deformable objects.
  arXiv  Detail & Related papers  (2024-07-21T16:15:02Z)
• Dynamic Handover: Throw and Catch with Bimanual Hands [30.206469112964033]
  We design a system with two multi-finger hands attached to robot arms to solve this problem. We train our system using Multi-Agent Reinforcement Learning in simulation and perform Sim2Real transfer to deploy on the real robots. To overcome the Sim2Real gap, we provide multiple novel algorithm designs including learning a trajectory prediction model for the object.
  arXiv  Detail & Related papers  (2023-09-11T17:49:25Z)
• Robot Learning with Sensorimotor Pre-training [98.7755895548928]
  We present a self-supervised sensorimotor pre-training approach for robotics. Our model, called RPT, is a Transformer that operates on sequences of sensorimotor tokens. We find that sensorimotor pre-training consistently outperforms training from scratch, has favorable scaling properties, and enables transfer across different tasks, environments, and robots.
  arXiv  Detail & Related papers  (2023-06-16T17:58:10Z)
• Malleable Agents for Re-Configurable Robotic Manipulators [0.0]
  We propose an RL agent with sequence neural networks embedded in the deep neural network to adapt to robotic arms with varying number of links. With the additional tool of domain randomization, this agent adapts to different configurations with varying number/length of links and dynamics noise.
  arXiv  Detail & Related papers  (2022-02-04T21:22:00Z)
• Towards Optimal Strategies for Training Self-Driving Perception Models in Simulation [98.51313127382937]
  We focus on the use of labels in the synthetic domain alone. Our approach introduces both a way to learn neural-invariant representations and a theoretically inspired view on how to sample the data from the simulator. We showcase our approach on the bird's-eye-view vehicle segmentation task with multi-sensor data.
  arXiv  Detail & Related papers  (2021-11-15T18:37:43Z)
• Robot Learning from Randomized Simulations: A Review [59.992761565399185]
  Deep learning has caused a paradigm shift in robotics research, favoring methods that require large amounts of data. State-of-the-art approaches learn in simulation where data generation is fast as well as inexpensive. We focus on a technique named 'domain randomization' which is a method for learning from randomized simulations.
  arXiv  Detail & Related papers  (2021-11-01T13:55:41Z)
• Cross Domain Robot Imitation with Invariant Representation [32.1735585546968]
  Cross domain imitation learning (CDIL) is a challenging task in robotics. We introduce an imitation learning algorithm based on invariant representation. We show that our method is able to learn similar representations for different robots with similar behaviors.
  arXiv  Detail & Related papers  (2021-09-13T13:05:35Z)
• Domain Adaptive Robotic Gesture Recognition with Unsupervised Kinematic-Visual Data Alignment [60.31418655784291]
  We propose a novel unsupervised domain adaptation framework which can simultaneously transfer multi-modality knowledge, i.e., both kinematic and visual data, from simulator to real robot. It remedies the domain gap with enhanced transferable features by using temporal cues in videos, and inherent correlations in multi-modal towards recognizing gesture. Results show that our approach recovers the performance with great improvement gains, up to 12.91% in ACC and 20.16% in F1score without using any annotations in real robot.
  arXiv  Detail & Related papers  (2021-03-06T09:10:03Z)
• Metric-Based Imitation Learning Between Two Dissimilar Anthropomorphic Robotic Arms [29.08134072341867]
  One major challenge in imitation learning is the correspondence problem. We introduce a distance measure between dissimilar embodiments. We find that the measure is well suited for describing the similarity between embodiments and for learning imitation policies by distance.
  arXiv  Detail & Related papers  (2020-02-25T19:47:19Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.