Trajectory adjustment for nonprehensile manipulation using latent space of trained sequence-to-sequence model. (2nd November 2019)

• Record Type: Journal Article
• Authors: Kutsuzawa, K.; Sakaino, S.; Tsuji, T.
• Is Part Of: Advanced robotics. Volume 33:Number 21(2019)
• Page Start: 1144
• Page End: 1154
• Publication Date: 2019-11-02
• Subjects: Nonprehensile manipulation -- sequence-to-sequence model -- latent variable -- neural network -- motion optimization; Robotics -- Periodicals; Robotics -- Japan -- Periodicals; Robotics; Japan; Periodicals; 629.89205
• DOI: 10.1080/01691864.2019.1673204
• Languages: English
• ISSNs: 0169-1864

Abstract

Abstract : When robots are used to manipulate objects in various ways, they often have to consider the dynamic constraint. Machine learning is a good candidate for such complex trajectory planning problems. However, it sometimes does not satisfy the task objectives due to a change in the objective or a lack of guarantee that the objective functions will be satisfied. To overcome this issue, we applied a method of trajectory deformation by using sequence-to-sequence (seq2seq) models. We propose a method of adjusting the generated trajectories, by utilizing the architecture of seq2seq models. The proposed method optimizes the latent variables of the seq2seq models instead of the trajectories to minimize the given objective functions. The verification results show that the use of latent variables can obtain the desired trajectories faster than direct optimization of the trajectories. GRAPHICAL ABSTRACT: