elib
DLR-Header
DLR-Logo -> http://www.dlr.de
DLR Portal Home | Imprint | Privacy Policy | Contact | Deutsch
Fontsize: [-] Text [+]

Mixture of experts on Riemannian manifolds for visual-servoing fixtures

Mühlbauer, Maximilian Sebastian and Stulp, Freek and Albu-Schäffer, Alin Olimpiu and Silvério, João (2022) Mixture of experts on Riemannian manifolds for visual-servoing fixtures. In: 2022 IEEE/RSJ International Conference on Intelligent Robots and Systems, Workshop on Probabilistic Robotics in the Age of Deep Learning. IROS 2022 Workshop Probabilistic Robotics in the Age of Deep Learning, 2022-10-27, Kyoto, Japan.

[img] PDF
3MB

Official URL: https://probabilisticrobotics.github.io/#04

Abstract

Adaptive Virtual Fixtures (VFs) for teleoperation often rely on visual inputs for online adaptation. State estimation from visual detections is never perfect, and thus affects the quality and robustness of adaptation. It is therefore important to be able to quantify how uncertain an estimation from vision is. This can, for example, inform on how to modulate a fixture's stiffness to decrease the physical force a human operator has to apply. Furthermore, the target of a manipulation operation might not be known from the beginning of the task, which creates the need for a principled way to add and remove fixtures when possible targets appear in the robot workspace. In this paper we propose an on-manifold Mixture of Experts (MoE) model that synthesizes visual-servoing fixtures while elegantly handling full pose detection uncertainties and 6D teleoperation goals in a unified framework. An arbitration function allocating the authority between multiple vision-based fixtures arises naturally from the MoE formulation. We show that this approach allows a teleoperator to insert multiple printed circuit boards (PCBs) with high precision without requiring the manual design of VFs to guide the robot motion. An exemplary video visualizing the probability distribution resulting from our model is available at: https://youtu.be/GKMQvbJ5OzA

Item URL in elib:https://elib.dlr.de/189970/
Document Type:Conference or Workshop Item (Poster)
Title:Mixture of experts on Riemannian manifolds for visual-servoing fixtures
Authors:
AuthorsInstitution or Email of AuthorsAuthor's ORCID iDORCID Put Code
Mühlbauer, Maximilian SebastianUNSPECIFIEDhttps://orcid.org/0000-0002-7635-0248UNSPECIFIED
Stulp, FreekUNSPECIFIEDhttps://orcid.org/0000-0001-9555-9517UNSPECIFIED
Albu-Schäffer, Alin OlimpiuUNSPECIFIEDhttps://orcid.org/0000-0001-5343-9074142115916
Silvério, JoãoUNSPECIFIEDhttps://orcid.org/0000-0003-1428-8933UNSPECIFIED
Date:27 October 2022
Journal or Publication Title:2022 IEEE/RSJ International Conference on Intelligent Robots and Systems, Workshop on Probabilistic Robotics in the Age of Deep Learning
Refereed publication:Yes
Open Access:Yes
Gold Open Access:No
In SCOPUS:No
In ISI Web of Science:No
Status:Published
Keywords:Mixture of experts, Riemannian manifolds, shared control, virtual fixtures
Event Title:IROS 2022 Workshop Probabilistic Robotics in the Age of Deep Learning
Event Location:Kyoto, Japan
Event Type:Workshop
Event Date:27 October 2022
HGF - Research field:Aeronautics, Space and Transport
HGF - Program:Space
HGF - Program Themes:Robotics
DLR - Research area:Raumfahrt
DLR - Program:R RO - Robotics
DLR - Research theme (Project):R - Telerobotics, R - Explainable Robotic AI
Location: Oberpfaffenhofen
Institutes and Institutions:Institute of Robotics and Mechatronics (since 2013) > Cognitive Robotics
Deposited By: Mühlbauer, Maximilian Sebastian
Deposited On:05 Dec 2022 14:36
Last Modified:24 Apr 2024 20:51

Repository Staff Only: item control page

Browse
Search
Help & Contact
Information
electronic library is running on EPrints 3.3.12
Website and database design: Copyright © German Aerospace Center (DLR). All rights reserved.