VoxAct-B: Voxel-Based Acting and Stabilizing Policy for Bimanual Manipulation

I-Chun Arthur Liu, Sicheng He, Daniel Seita, Gaurav S. Sukhatme
Proceedings of The 8th Conference on Robot Learning, PMLR 270:4354-4370, 2025.

Abstract

Bimanual manipulation is critical to many robotics applications. In contrast to single-arm manipulation, bimanual manipulation tasks are challenging due to higher-dimensional action spaces. Prior works leverage large amounts of data and primitive actions to address this problem, but may suffer from sample inefficiency and limited generalization across various tasks. To this end, we propose VoxAct-B, a language-conditioned, voxel-based method that leverages Vision Language Models (VLMs) to prioritize key regions within the scene and reconstruct a voxel grid. We provide this voxel grid to our bimanual manipulation policy to learn acting and stabilizing actions. This approach enables more efficient policy learning from voxels and is generalizable to different tasks. In simulation, we show that VoxAct-B outperforms strong baselines on fine-grained bimanual manipulation tasks. Furthermore, we demonstrate VoxAct-B on real-world Open Drawer and Open Jar tasks using two UR5s. Code, data, and videos are available at https://voxact-b.github.io.

Cite this Paper


BibTeX
@InProceedings{pmlr-v270-liu25i, title = {VoxAct-B: Voxel-Based Acting and Stabilizing Policy for Bimanual Manipulation}, author = {Liu, I-Chun Arthur and He, Sicheng and Seita, Daniel and Sukhatme, Gaurav S.}, booktitle = {Proceedings of The 8th Conference on Robot Learning}, pages = {4354--4370}, year = {2025}, editor = {Agrawal, Pulkit and Kroemer, Oliver and Burgard, Wolfram}, volume = {270}, series = {Proceedings of Machine Learning Research}, month = {06--09 Nov}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v270/main/assets/liu25i/liu25i.pdf}, url = {https://proceedings.mlr.press/v270/liu25i.html}, abstract = {Bimanual manipulation is critical to many robotics applications. In contrast to single-arm manipulation, bimanual manipulation tasks are challenging due to higher-dimensional action spaces. Prior works leverage large amounts of data and primitive actions to address this problem, but may suffer from sample inefficiency and limited generalization across various tasks. To this end, we propose VoxAct-B, a language-conditioned, voxel-based method that leverages Vision Language Models (VLMs) to prioritize key regions within the scene and reconstruct a voxel grid. We provide this voxel grid to our bimanual manipulation policy to learn acting and stabilizing actions. This approach enables more efficient policy learning from voxels and is generalizable to different tasks. In simulation, we show that VoxAct-B outperforms strong baselines on fine-grained bimanual manipulation tasks. Furthermore, we demonstrate VoxAct-B on real-world $\texttt{Open Drawer}$ and $\texttt{Open Jar}$ tasks using two UR5s. Code, data, and videos are available at https://voxact-b.github.io.} }
Endnote
%0 Conference Paper %T VoxAct-B: Voxel-Based Acting and Stabilizing Policy for Bimanual Manipulation %A I-Chun Arthur Liu %A Sicheng He %A Daniel Seita %A Gaurav S. Sukhatme %B Proceedings of The 8th Conference on Robot Learning %C Proceedings of Machine Learning Research %D 2025 %E Pulkit Agrawal %E Oliver Kroemer %E Wolfram Burgard %F pmlr-v270-liu25i %I PMLR %P 4354--4370 %U https://proceedings.mlr.press/v270/liu25i.html %V 270 %X Bimanual manipulation is critical to many robotics applications. In contrast to single-arm manipulation, bimanual manipulation tasks are challenging due to higher-dimensional action spaces. Prior works leverage large amounts of data and primitive actions to address this problem, but may suffer from sample inefficiency and limited generalization across various tasks. To this end, we propose VoxAct-B, a language-conditioned, voxel-based method that leverages Vision Language Models (VLMs) to prioritize key regions within the scene and reconstruct a voxel grid. We provide this voxel grid to our bimanual manipulation policy to learn acting and stabilizing actions. This approach enables more efficient policy learning from voxels and is generalizable to different tasks. In simulation, we show that VoxAct-B outperforms strong baselines on fine-grained bimanual manipulation tasks. Furthermore, we demonstrate VoxAct-B on real-world $\texttt{Open Drawer}$ and $\texttt{Open Jar}$ tasks using two UR5s. Code, data, and videos are available at https://voxact-b.github.io.
APA
Liu, I.A., He, S., Seita, D. & Sukhatme, G.S.. (2025). VoxAct-B: Voxel-Based Acting and Stabilizing Policy for Bimanual Manipulation. Proceedings of The 8th Conference on Robot Learning, in Proceedings of Machine Learning Research 270:4354-4370 Available from https://proceedings.mlr.press/v270/liu25i.html.

Related Material