Mixtures of Controlled Gaussian Processes for Dynamical Modeling of Deformable Objects

Ce Xu Zheng, Adriá Colomé, Luis Sentis, Carme Torras
Proceedings of The 4th Annual Learning for Dynamics and Control Conference, PMLR 168:415-426, 2022.

Abstract

Control and manipulation of objects is a highly relevant topic in Robotics research. Although significant advances have been made over the manipulation of rigid bodies, the manipulation of non-rigid objects is still challenging and an open problem. Due to the uncertainty of the outcome when applying physical actions to non-rigid objects, using prior knowledge on objects’ dynamics can greatly improve the control performance. However, fitting such models is a challenging task for materials such as clothing, where the state is represented by points in a mesh, resulting in very large dimensionality that makes models difficult to learn, process and predict based on measured data. In this paper, we expand previous work on Controlled Gaussian Process Dynamical Models (CGPDM), a method that uses a non-linear projection of the state space onto a much smaller dimensional latent space, and learns the object dynamics in the latent space. We take advantage of the variability in training data by employing Mixture of Experts (MoE), and we devise theory and experimental validations that demonstrate significant improvements in training and prediction times, plus robustness and error stability when predicting deformable objects exposed to disparate movement ranges.

Cite this Paper


BibTeX
@InProceedings{pmlr-v168-zheng22a, title = {Mixtures of Controlled Gaussian Processes for Dynamical Modeling of Deformable Objects}, author = {Zheng, Ce Xu and Colom\'e, Adri\'a and Sentis, Luis and Torras, Carme}, booktitle = {Proceedings of The 4th Annual Learning for Dynamics and Control Conference}, pages = {415--426}, year = {2022}, editor = {Firoozi, Roya and Mehr, Negar and Yel, Esen and Antonova, Rika and Bohg, Jeannette and Schwager, Mac and Kochenderfer, Mykel}, volume = {168}, series = {Proceedings of Machine Learning Research}, month = {23--24 Jun}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v168/zheng22a/zheng22a.pdf}, url = {https://proceedings.mlr.press/v168/zheng22a.html}, abstract = {Control and manipulation of objects is a highly relevant topic in Robotics research. Although significant advances have been made over the manipulation of rigid bodies, the manipulation of non-rigid objects is still challenging and an open problem. Due to the uncertainty of the outcome when applying physical actions to non-rigid objects, using prior knowledge on objects’ dynamics can greatly improve the control performance. However, fitting such models is a challenging task for materials such as clothing, where the state is represented by points in a mesh, resulting in very large dimensionality that makes models difficult to learn, process and predict based on measured data. In this paper, we expand previous work on Controlled Gaussian Process Dynamical Models (CGPDM), a method that uses a non-linear projection of the state space onto a much smaller dimensional latent space, and learns the object dynamics in the latent space. We take advantage of the variability in training data by employing Mixture of Experts (MoE), and we devise theory and experimental validations that demonstrate significant improvements in training and prediction times, plus robustness and error stability when predicting deformable objects exposed to disparate movement ranges.} }
Endnote
%0 Conference Paper %T Mixtures of Controlled Gaussian Processes for Dynamical Modeling of Deformable Objects %A Ce Xu Zheng %A Adriá Colomé %A Luis Sentis %A Carme Torras %B Proceedings of The 4th Annual Learning for Dynamics and Control Conference %C Proceedings of Machine Learning Research %D 2022 %E Roya Firoozi %E Negar Mehr %E Esen Yel %E Rika Antonova %E Jeannette Bohg %E Mac Schwager %E Mykel Kochenderfer %F pmlr-v168-zheng22a %I PMLR %P 415--426 %U https://proceedings.mlr.press/v168/zheng22a.html %V 168 %X Control and manipulation of objects is a highly relevant topic in Robotics research. Although significant advances have been made over the manipulation of rigid bodies, the manipulation of non-rigid objects is still challenging and an open problem. Due to the uncertainty of the outcome when applying physical actions to non-rigid objects, using prior knowledge on objects’ dynamics can greatly improve the control performance. However, fitting such models is a challenging task for materials such as clothing, where the state is represented by points in a mesh, resulting in very large dimensionality that makes models difficult to learn, process and predict based on measured data. In this paper, we expand previous work on Controlled Gaussian Process Dynamical Models (CGPDM), a method that uses a non-linear projection of the state space onto a much smaller dimensional latent space, and learns the object dynamics in the latent space. We take advantage of the variability in training data by employing Mixture of Experts (MoE), and we devise theory and experimental validations that demonstrate significant improvements in training and prediction times, plus robustness and error stability when predicting deformable objects exposed to disparate movement ranges.
APA
Zheng, C.X., Colomé, A., Sentis, L. & Torras, C.. (2022). Mixtures of Controlled Gaussian Processes for Dynamical Modeling of Deformable Objects. Proceedings of The 4th Annual Learning for Dynamics and Control Conference, in Proceedings of Machine Learning Research 168:415-426 Available from https://proceedings.mlr.press/v168/zheng22a.html.

Related Material