Deep Imitation Learning for Bimanual Robotic Manipulation

Abstract

We present a deep imitation learning framework for robotic bimanual manipulation in a continuous state-action space. A core challenge is to generalize the manipulation skills to objects in different locations. We hypothesize that modeling the relational information in the environment can significantly improve generalization. To achieve this, we propose to (i) decompose the multi-modal dynamics into elemental movement primitives, (ii) parameterize each primitive using a recurrent graph neural network to capture interactions, and (iii) integrate a high-level planner that composes primitives sequentially and a low-level controller to combine primitive dynamics and inverse kinematics control. Our model is a deep, hierarchical, modular architecture. Compared to baselines, our model generalizes better and achieves higher success rates on several simulated bimanual robotic manipulation tasks. We open source the code for simulation, data, and models at: https://github.com/Rose-STL-Lab/HDR-IL.

References in corpus (10)

• Sequence to Sequence Learning with Neural Networks
• Neural Relational Inference for Interacting Systems
• Parametrized Deep Q-Networks Learning: Reinforcement Learning with Discrete-Continuous Hybrid Action Space
• Learning Parameterized Skills
• Reinforcement Learning with Parameterized Actions
• Deep Imitative Models for Flexible Inference, Planning, and Control
• One-Shot Hierarchical Imitation Learning of Compound Visuomotor Tasks
• Dynamics Learning with Cascaded Variational Inference for Multi-Step Manipulation
• Object-Oriented Dynamics Predictor
• Hierarchical Variational Imitation Learning of Control Programs

Cited by in corpus (4)

• Imitation and Mirror Systems in Robots through Deep Modality Blending Networks
• Cloud-Based Hierarchical Imitation Learning for Scalable Transfer of Construction Skills from Human Workers to Assisting Robots
• DAIR: Disentangled Attention Intrinsic Regularization for Safe and Efficient Bimanual Manipulation
• Wish you were here: Hindsight Goal Selection for long-horizon dexterous manipulation