Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Video

Updated: Jul 9, 2025

Estimation of Contact Regions Between Hands and Objects During Human Multi-Digit Grasping
09:41

Estimation of Contact Regions Between Hands and Objects During Human Multi-Digit Grasping

Published on: April 21, 2023

1.6K

Peg-in-hole assembly skill imitation learning method based on ProMPs under task geometric representation.

Yajing Zang1, Pengfei Wang1, Fusheng Zha1

  • 1School of Mechatronics Engineering, State Key Laboratory of Robotics and System, Harbin Institute of Technology, Harbin, China.

Frontiers in Neurorobotics
|November 29, 2023
PubMed
Summary

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Quantum Simulation of Pseudo-Hermitian-<i>φ</i>-Symmetric Two-Level Systems.

Entropy (Basel, Switzerland)·2022
Same author

Unexpected toroidal micelles formed from St/MMA gradient copolymers.

Soft matter·2022
Same author

Characterization of <i>Pseudomonas aeruginosa</i> Bacteriophage L5 Which Requires Type IV Pili for Infection.

Frontiers in microbiology·2022
Same author

Constructing high-performance TADF polymers from non-TADF monomers: a computational investigation.

Physical chemistry chemical physics : PCCP·2022
Same author

Research on the Mechanical Properties and Stretch Forming Simulation of Triaxial Geogrid with Different Pre-Punched Hole Diameters.

Polymers·2022
Same author

Spatial distribution of soil nutrients and evaluation of cultivated land in Xuwen county.

PeerJ·2022

This study introduces a streamlined imitation learning method using geometric representations and Probability Movement Primitives (ProMPs) to improve robotic assembly task learning from sparse human demonstrations. The approach enhances Behavioral Cloning (BC) for better skill acquisition and generalization.

Area of Science:

  • Robotics
  • Machine Learning
  • Imitation Learning

Background:

  • Behavioral Cloning (BC) struggles with sparse and imperfect human demonstration data for robotic manipulation.
  • Existing methods face challenges in effectively utilizing limited demonstration trajectories for complex assembly tasks.

Purpose of the Study:

  • To propose a streamlined imitation learning method that leverages geometric representations for enhanced robotic assembly skill learning.
  • To improve the utilization of sparse and imperfect human demonstration data in robotic manipulation tasks.

Main Methods:

  • Demonstration trajectories are mapped to a geometric feature space and aligned using Dynamic Time Warping (DTW).
  • Probability Movement Primitives (ProMPs) are extracted to generate diverse task trajectories for training neural networks via Behavioral Cloning (BC).
Keywords:
Behavioral Cloningimitation learningpeg-in-hole assemblyprobabilistic movement primitivesrobot manipulation planning

More Related Videos

Investigating Motor Skill Learning Processes with a Robotic Manipulandum
07:52

Investigating Motor Skill Learning Processes with a Robotic Manipulandum

Published on: February 12, 2017

8.8K
Study Motor Skill Learning by Single-pellet Reaching Tasks in Mice
06:04

Study Motor Skill Learning by Single-pellet Reaching Tasks in Mice

Published on: March 4, 2014

21.0K

Related Experiment Videos

Last Updated: Jul 9, 2025

Estimation of Contact Regions Between Hands and Objects During Human Multi-Digit Grasping
09:41

Estimation of Contact Regions Between Hands and Objects During Human Multi-Digit Grasping

Published on: April 21, 2023

1.6K
Investigating Motor Skill Learning Processes with a Robotic Manipulandum
07:52

Investigating Motor Skill Learning Processes with a Robotic Manipulandum

Published on: February 12, 2017

8.8K
Study Motor Skill Learning by Single-pellet Reaching Tasks in Mice
06:04

Study Motor Skill Learning by Single-pellet Reaching Tasks in Mice

Published on: March 4, 2014

21.0K
  • The method utilizes current task states as via points in the ProMPs model to generate future trajectories.

    • Main Results:

    • The proposed method achieved higher success rates in peg-in-hole assembly tasks compared to traditional imitation learning.
    • The learned assembly strategy demonstrated reasonable generalization capabilities in simulation and on a real robotic platform.
    • Geometric representation with ProMPs effectively improved BC's ability to learn task skills from demonstration data.

    Conclusions:

    • The streamlined imitation learning method enhances BC by effectively utilizing geometric representations and ProMPs.
    • This approach offers a viable solution for learning complex robotic manipulation skills from limited human demonstrations.
    • The findings highlight the potential of geometric feature extraction and ProMPs for improving imitation learning in robotics.