Computer Science > Robotics
[Submitted on 1 Sep 2023]
Title:Deep Segmented DMP Networks for Learning Discontinuous Motions
View PDFAbstract:Discontinuous motion which is a motion composed of multiple continuous motions with sudden change in direction or velocity in between, can be seen in state-aware robotic tasks. Such robotic tasks are often coordinated with sensor information such as image. In recent years, Dynamic Movement Primitives (DMP) which is a method for generating motor behaviors suitable for robotics has garnered several deep learning based improvements to allow associations between sensor information and DMP parameters. While the implementation of deep learning framework does improve upon DMP's inability to directly associate to an input, we found that it has difficulty learning DMP parameters for complex motion which requires large number of basis functions to reconstruct. In this paper we propose a novel deep learning network architecture called Deep Segmented DMP Network (DSDNet) which generates variable-length segmented motion by utilizing the combination of multiple DMP parameters predicting network architecture, double-stage decoder network, and number of segments predictor. The proposed method is evaluated on both artificial data (object cutting & pick-and-place) and real data (object cutting) where our proposed method could achieve high generalization capability, task-achievement, and data-efficiency compared to previous method on generating discontinuous long-horizon motions.
References & Citations
Bibliographic and Citation Tools
Bibliographic Explorer (What is the Explorer?)
Litmaps (What is Litmaps?)
scite Smart Citations (What are Smart Citations?)
Code, Data and Media Associated with this Article
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub (What is DagsHub?)
Gotit.pub (What is GotitPub?)
Papers with Code (What is Papers with Code?)
ScienceCast (What is ScienceCast?)
Demos
Recommenders and Search Tools
Influence Flower (What are Influence Flowers?)
Connected Papers (What is Connected Papers?)
CORE Recommender (What is CORE?)
arXivLabs: experimental projects with community collaborators
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.