Mixtures of controlled Gaussian processes for dynamical modeling of deformable objects
Tipus de documentText en actes de congrés
Data publicació2022
EditorProceedings of Machine Learning Research (PMLR)
Condicions d'accésAccés obert
Tots els drets reservats. Aquesta obra està protegida pels drets de propietat intel·lectual i
industrial corresponents. Sense perjudici de les exempcions legals existents, queda prohibida la seva
reproducció, distribució, comunicació pública o transformació sense l'autorització del titular dels drets
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.
CitacióXu, C. [et al.]. Mixtures of controlled Gaussian processes for dynamical modeling of deformable objects. A: Annual Learning for Dynamics & Control Conference. "Proceedings of The 4th Annual Learning for Dynamics and Control Conference. Volume 168: Learning for Dynamics and Control Conference, 23-24 June 2022, Stanford University, Stanford, CA, USA". Proceedings of Machine Learning Research (PMLR), 2022, p. 415-426. ISBN 2640-3498.
ISBN2640-3498
Versió de l'editorhttps://proceedings.mlr.press/v168/zheng22a.html
Col·leccions
Fitxers | Descripció | Mida | Format | Visualitza |
---|---|---|---|---|
2593-Mixtures-o ... -Deformable-Objects(1).pdf | 288,4Kb | Visualitza/Obre |