Cross Domain Robot Imitation with Invariant Representation

Yin, Zhao-Heng; Sun, Lingfeng; Ma, Hengbo; Tomizuka, Masayoshi; Li, Wu-Jun

doi:10.1109/icra46639.2022.9811668

Cited by 4 publications

(2 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Transfer learning has shown good performance to share prior knowledge to improve the efficiency of training and generalization ability in robotics and reinforcement learning [11], [12]. In [13], transferring end-to-end controllers to the real world is accomplished by using demonstrations of linear paths constructed via inverse kinematics (IK) in Cartesian space, to construct a dataset that can then be used to train a reactive neural network controller which continuously accepts images along with joint angles, and outputs motor velocities.…”

Section: A Transfer Learning In Roboticsmentioning

confidence: 99%

Learn from Robot: Transferring Skills for Diverse Manipulation via Cycle Generative Networks

Yang,

Stork,

Stoyanov

2023

2023 IEEE 19th International Conference on Automation Science and Engineering (CASE)

View full text Add to dashboard Cite

Reinforcement learning (RL) has shown impressive results on a variety of robot tasks, but it requires a large amount of data for learning a single RL policy. However, in manufacturing there is a wide demand of reusing skills from different robots and it is hard to transfer the learned policy to different hardware due to diverse robot body morphology, kinematics, and dynamics. In this paper, we address the problem of transferring policies between different robot platforms. We learn a set of skills on each specific robot and represent them in a latent space. We propose to transfer the skills between different robots by mapping latent action spaces through a cycle generative network in a supervised learning manner. We extend the policy model learned on one robot with a pre-trained generative network to enable the robot to learn from the skill of another robot. We evaluate our method on several simulated experiments and demonstrate that our Learn from Robot (LfR) method accelerates new skill learning.

show abstract

Section: A Transfer Learning In Roboticsmentioning

confidence: 99%

Learn from Robot: Transferring Skills for Diverse Manipulation via Cycle Generative Networks

Yang,

Stork,

Stoyanov

2023

2023 IEEE 19th International Conference on Automation Science and Engineering (CASE)

View full text Add to dashboard Cite

show abstract

“…In contrast, our work does not require expert demonstrations in target domains. Yin et al [32] learn a latent invariant representation for a robot with different physical parameters (e.g. link length).…”

Section: Related Workmentioning

confidence: 99%

Learning Latent Object-Centric Representations for Visual-Based Robot Manipulation

Wang

et al. 2022

2022 International Conference on Advanced Robotics and Mechatronics (ICARM)

View full text Add to dashboard Cite

This paper focuses on transferring control policies between robot manipulators with different morphology. While reinforcement learning (RL) methods have shown successful results in robot manipulation tasks, transferring a trained policy from simulation to a real robot or deploying it on a robot with different states, actions, or kinematics is challenging. To achieve cross-embodiment policy transfer, our key insight is to project the state and action spaces of the source and target robots to a common latent space representation. We first introduce encoders and decoders to associate the states and actions of the source robot with a latent space. The encoders, decoders, and a latent space control policy are trained simultaneously using loss functions measuring task performance, latent dynamics consistency, and encoder-decoder ability to reconstruct the original states and actions. To transfer the learned control policy, we only need to train target encoders and decoders that align a new target domain to the latent space. We use generative adversarial training with cycle consistency and latent dynamics losses without access to the task reward or reward tuning in the target domain. We demonstrate sim-to-sim and sim-to-real manipulation policy transfer with source and target robots of different states, actions, and embodiments. The source code is available at https://github.com/ExistentialRobotics/ cross_embodiment_transfer.

show abstract