OrbitGrasp: SE(3)-Equivariant Grasp Learning

Boce Hu, Xupeng Zhu, Dian Wang, Zihao Dong, Haojie Huang, Chenghao Wang, Robin Walters, Robert Platt
Proceedings of The 8th Conference on Robot Learning, PMLR 270:2456-2474, 2025.

Abstract

While grasp detection is an important part of any robotic manipulation pipeline, reliable and accurate grasp detection in mathrmSE(3) remains a research challenge. Many robotics applications in unstructured environments such as the home or warehouse would benefit a lot from better grasp performance. This paper proposes a novel framework for detecting SE(3) grasp poses based on point cloud input. Our main contribution is to propose an SE(3)-equivariant model that maps each point in the cloud to a continuous grasp quality function over the 2-sphere S2 using a spherical harmonic basis. Compared with reasoning about a finite set of samples, this formulation improves the accuracy and efficiency of our model when a large number of samples would otherwise be needed. In order to accomplish this, we propose a novel variation on EquiFormerV2 that leverages a UNet-style backbone to enlarge the number of points the model can handle. Our resulting method, which we name OrbitGrasp, significantly outperforms baselines in both simulation and physical experiments.

Cite this Paper

BibTeX
@InProceedings{pmlr-v270-hu25b, title = {OrbitGrasp: SE(3)-Equivariant Grasp Learning}, author = {Hu, Boce and Zhu, Xupeng and Wang, Dian and Dong, Zihao and Huang, Haojie and Wang, Chenghao and Walters, Robin and Platt, Robert}, booktitle = {Proceedings of The 8th Conference on Robot Learning}, pages = {2456--2474}, 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/hu25b/hu25b.pdf}, url = {https://proceedings.mlr.press/v270/hu25b.html}, abstract = {While grasp detection is an important part of any robotic manipulation pipeline, reliable and accurate grasp detection in $\\mathrm{SE}(3)$ remains a research challenge. Many robotics applications in unstructured environments such as the home or warehouse would benefit a lot from better grasp performance. This paper proposes a novel framework for detecting $\mathrm{SE}(3)$ grasp poses based on point cloud input. Our main contribution is to propose an $\mathrm{SE}(3)$-equivariant model that maps each point in the cloud to a continuous grasp quality function over the 2-sphere $S^2$ using a spherical harmonic basis. Compared with reasoning about a finite set of samples, this formulation improves the accuracy and efficiency of our model when a large number of samples would otherwise be needed. In order to accomplish this, we propose a novel variation on EquiFormerV2 that leverages a UNet-style backbone to enlarge the number of points the model can handle. Our resulting method, which we name OrbitGrasp, significantly outperforms baselines in both simulation and physical experiments.} }
Endnote
%0 Conference Paper %T OrbitGrasp: SE(3)-Equivariant Grasp Learning %A Boce Hu %A Xupeng Zhu %A Dian Wang %A Zihao Dong %A Haojie Huang %A Chenghao Wang %A Robin Walters %A Robert Platt %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-hu25b %I PMLR %P 2456--2474 %U https://proceedings.mlr.press/v270/hu25b.html %V 270 %X While grasp detection is an important part of any robotic manipulation pipeline, reliable and accurate grasp detection in $\\mathrm{SE}(3)$ remains a research challenge. Many robotics applications in unstructured environments such as the home or warehouse would benefit a lot from better grasp performance. This paper proposes a novel framework for detecting $\mathrm{SE}(3)$ grasp poses based on point cloud input. Our main contribution is to propose an $\mathrm{SE}(3)$-equivariant model that maps each point in the cloud to a continuous grasp quality function over the 2-sphere $S^2$ using a spherical harmonic basis. Compared with reasoning about a finite set of samples, this formulation improves the accuracy and efficiency of our model when a large number of samples would otherwise be needed. In order to accomplish this, we propose a novel variation on EquiFormerV2 that leverages a UNet-style backbone to enlarge the number of points the model can handle. Our resulting method, which we name OrbitGrasp, significantly outperforms baselines in both simulation and physical experiments.
APA
Hu, B., Zhu, X., Wang, D., Dong, Z., Huang, H., Wang, C., Walters, R. & Platt, R.. (2025). OrbitGrasp: SE(3)-Equivariant Grasp Learning. Proceedings of The 8th Conference on Robot Learning, in Proceedings of Machine Learning Research 270:2456-2474 Available from https://proceedings.mlr.press/v270/hu25b.html.

Related Material