Fast Flow-based Visuomotor Policies via Conditional Optimal Transport Couplings

Andreas Sochopoulos, Nikolay Malkin, Nikolaos Tsagkas, Joao Moura, Michael Gienger, Sethu Vijayakumar
Proceedings of The 9th Conference on Robot Learning, PMLR 305:3357-3377, 2025.

Abstract

Diffusion and flow matching policies have recently demonstrated remarkable performance in robotic applications by accurately capturing multimodal robot trajectory distributions. However, their computationally expensive inference, due to the numerical integration of an ODE or SDE, limits their applicability as real-time controllers for robots. We introduce a methodology that utilizes conditional Optimal Transport couplings between noise and samples to enforce straight solutions in the flow ODE for robot action generation tasks. We show that naively coupling noise and samples fails in conditional tasks and propose incorporating condition variables into the coupling process to improve few-step performance. The proposed few-step policy achieves a 4% higher success rate with a 10$\times$ speed-up compared to Diffusion Policy on a diverse set of simulation tasks. Moreover, it produces high-quality and diverse action trajectories within 1-2 steps on a set of real-world robot tasks. Our method also retains the same training complexity as Diffusion Policy and vanilla Flow Matching, in contrast to distillation-based approaches.

Cite this Paper

BibTeX
@InProceedings{pmlr-v305-sochopoulos25a, title = {Fast Flow-based Visuomotor Policies via Conditional Optimal Transport Couplings}, author = {Sochopoulos, Andreas and Malkin, Nikolay and Tsagkas, Nikolaos and Moura, Joao and Gienger, Michael and Vijayakumar, Sethu}, booktitle = {Proceedings of The 9th Conference on Robot Learning}, pages = {3357--3377}, year = {2025}, editor = {Lim, Joseph and Song, Shuran and Park, Hae-Won}, volume = {305}, series = {Proceedings of Machine Learning Research}, month = {27--30 Sep}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v305/main/assets/sochopoulos25a/sochopoulos25a.pdf}, url = {https://proceedings.mlr.press/v305/sochopoulos25a.html}, abstract = {Diffusion and flow matching policies have recently demonstrated remarkable performance in robotic applications by accurately capturing multimodal robot trajectory distributions. However, their computationally expensive inference, due to the numerical integration of an ODE or SDE, limits their applicability as real-time controllers for robots. We introduce a methodology that utilizes conditional Optimal Transport couplings between noise and samples to enforce straight solutions in the flow ODE for robot action generation tasks. We show that naively coupling noise and samples fails in conditional tasks and propose incorporating condition variables into the coupling process to improve few-step performance. The proposed few-step policy achieves a 4% higher success rate with a 10$\times$ speed-up compared to Diffusion Policy on a diverse set of simulation tasks. Moreover, it produces high-quality and diverse action trajectories within 1-2 steps on a set of real-world robot tasks. Our method also retains the same training complexity as Diffusion Policy and vanilla Flow Matching, in contrast to distillation-based approaches.} }
Endnote
%0 Conference Paper %T Fast Flow-based Visuomotor Policies via Conditional Optimal Transport Couplings %A Andreas Sochopoulos %A Nikolay Malkin %A Nikolaos Tsagkas %A Joao Moura %A Michael Gienger %A Sethu Vijayakumar %B Proceedings of The 9th Conference on Robot Learning %C Proceedings of Machine Learning Research %D 2025 %E Joseph Lim %E Shuran Song %E Hae-Won Park %F pmlr-v305-sochopoulos25a %I PMLR %P 3357--3377 %U https://proceedings.mlr.press/v305/sochopoulos25a.html %V 305 %X Diffusion and flow matching policies have recently demonstrated remarkable performance in robotic applications by accurately capturing multimodal robot trajectory distributions. However, their computationally expensive inference, due to the numerical integration of an ODE or SDE, limits their applicability as real-time controllers for robots. We introduce a methodology that utilizes conditional Optimal Transport couplings between noise and samples to enforce straight solutions in the flow ODE for robot action generation tasks. We show that naively coupling noise and samples fails in conditional tasks and propose incorporating condition variables into the coupling process to improve few-step performance. The proposed few-step policy achieves a 4% higher success rate with a 10$\times$ speed-up compared to Diffusion Policy on a diverse set of simulation tasks. Moreover, it produces high-quality and diverse action trajectories within 1-2 steps on a set of real-world robot tasks. Our method also retains the same training complexity as Diffusion Policy and vanilla Flow Matching, in contrast to distillation-based approaches.
APA
Sochopoulos, A., Malkin, N., Tsagkas, N., Moura, J., Gienger, M. & Vijayakumar, S.. (2025). Fast Flow-based Visuomotor Policies via Conditional Optimal Transport Couplings. Proceedings of The 9th Conference on Robot Learning, in Proceedings of Machine Learning Research 305:3357-3377 Available from https://proceedings.mlr.press/v305/sochopoulos25a.html.

Related Material