AnyBimanual: Transferring Unimanual Policy for General Bimanual Manipulation

TL;DR

AnyBimanual is a plug-and-play approach that transfers pre-trained unimanual policies to bimanual manipulation tasks, significantly reducing data requirements and improving success rates in both simulated and real-world environments.

Contribution

The paper introduces a novel method combining skill management and visual alignment to effectively transfer unimanual policies to bimanual tasks with minimal demonstrations.

Findings

Achieved 12.67% higher success rate on simulated tasks.

Demonstrated 84.62% average success rate on real-world tasks.

Outperformed previous methods in general bimanual manipulation.

Abstract

Performing general language-conditioned bimanual manipulation tasks is of great importance for many applications ranging from household service to industrial assembly. However, collecting bimanual manipulation data is expensive due to the high-dimensional action space, which poses challenges for conventional methods to handle general bimanual manipulation tasks. In contrast, unimanual policy has recently demonstrated impressive generalizability across a wide range of tasks because of scaled model parameters and training data, which can provide sharable manipulation knowledge for bimanual systems. To this end, we propose a plug-and-play method named AnyBimanual, which transfers pre-trained unimanual policy to general bimanual manipulation policy with few bimanual demonstrations. Specifically, we first introduce a skill manager to dynamically schedule the skill representations discovered from pre-trained unimanual policy for bimanual manipulation tasks, which linearly combines skill primitives with task-oriented compensation to represent the bimanual manipulation instruction. To mitigate the observation discrepancy between unimanual and bimanual systems, we present a visual aligner to generate soft masks for visual embedding of the workspace, which aims to align visual input of unimanual policy model for each arm with those during pretraining stage. AnyBimanual shows superiority on 12 simulated tasks from RLBench2 with a sizable 12.67% improvement in success rate over previous methods. Experiments on 9 real-world tasks further verify its practicality with an average success rate of 84.62%.