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 Concept Videos

You might also read

Related Articles

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

Sort by
Same author

NURSE-AI: A Nurse-by-Design Framework for Multi-Sensor, AI-Enabled Chronic Wound Assessment in Community Healthcare.

Sensors (Basel, Switzerland)·2026
Same author

An evolutionary conserved neural mechanism for interpersonal coordination in primates.

Science advances·2026
Same author

Reinforcement Learning-Based Management in IoT-Enabled Renewable Energy Communities: An Approach to Optimization for Comfort, Economy, and Sustainable Performance.

Sensors (Basel, Switzerland)·2026
Same author

AI-Enhanced Hybrid QAM-PPM Visible Light Communication for Body Area Networks.

Sensors (Basel, Switzerland)·2026
Same author

Physiological Consequences of Overexpression of a Twin-Arginine Translocase in <i>Bacillus subtilis</i> Revealed by <sup>14</sup>N/<sup>15</sup>N Labeling.

Journal of proteome research·2025
Same author

Generalizable Hybrid Wavelet-Deep Learning Architecture for Robust Arrhythmia Detection in Wearable ECG Monitoring.

Sensors (Basel, Switzerland)·2025

Related Experiment Video

Updated: Jun 1, 2025

3D Kinematic Gait Analysis for Preclinical Studies in Rodents
10:19

3D Kinematic Gait Analysis for Preclinical Studies in Rodents

Published on: August 3, 2019

10.6K

DeepLabCut custom-trained model and the refinement function for gait analysis.

Giulia Panconi1, Stefano Grasso2, Sara Guarducci3

  • 1Department of Experimental and Clinical Medicine, University of Florence, Florence, Italy. giulia.panconi@unifi.it.

Scientific Reports
|January 17, 2025
PubMed
Summary

Markerless pose estimation using DeepLabCut (DLC) with custom training significantly improves human locomotion analysis compared to OpenPose (OP). DLC offers a promising, accurate, and low-cost solution for movement assessment outside the lab.

Keywords:
Deep learningDeepLabCutGait AnalysisOpenPosePose estimationVideo analysis

More Related Videos

Using Gold-standard Gait Analysis Methods to Assess Experience Effects on Lower-limb Mechanics During Moderate High-heeled Jogging and Running
06:35

Using Gold-standard Gait Analysis Methods to Assess Experience Effects on Lower-limb Mechanics During Moderate High-heeled Jogging and Running

Published on: September 14, 2017

9.1K
Clinical-oriented Three-dimensional Gait Analysis Method for Evaluating Gait Disorder
06:54

Clinical-oriented Three-dimensional Gait Analysis Method for Evaluating Gait Disorder

Published on: March 4, 2018

14.0K

Related Experiment Videos

Last Updated: Jun 1, 2025

3D Kinematic Gait Analysis for Preclinical Studies in Rodents
10:19

3D Kinematic Gait Analysis for Preclinical Studies in Rodents

Published on: August 3, 2019

10.6K
Using Gold-standard Gait Analysis Methods to Assess Experience Effects on Lower-limb Mechanics During Moderate High-heeled Jogging and Running
06:35

Using Gold-standard Gait Analysis Methods to Assess Experience Effects on Lower-limb Mechanics During Moderate High-heeled Jogging and Running

Published on: September 14, 2017

9.1K
Clinical-oriented Three-dimensional Gait Analysis Method for Evaluating Gait Disorder
06:54

Clinical-oriented Three-dimensional Gait Analysis Method for Evaluating Gait Disorder

Published on: March 4, 2018

14.0K

Area of Science:

  • Biomechanics
  • Computer Vision
  • Human Movement Analysis

Background:

  • Marker-based motion capture is precise but costly and environment-restricted.
  • Markerless pose estimation offers ecological, unobtrusive human movement data acquisition.
  • OpenPose (OP) and DeepLabCut (DLC) are popular markerless systems for movement analysis.

Purpose of the Study:

  • To compare the performance of OpenPose and DeepLabCut markerless systems for human locomotion assessment.
  • To evaluate the effectiveness of custom training and refinement functions in DeepLabCut for gait analysis.
  • To identify accurate, low-cost alternatives to traditional motion capture for movement studies.

Main Methods:

  • Forty healthy subjects walked on a 5m walkway with force platforms and a camera.
  • Gait parameters were collected using OpenPose (OPPT), pre-trained DeepLabCut (DLCPT), and custom-trained DeepLabCut (DLCCT).
  • Results were validated against force platform data as the reference system.

Main Results:

  • Custom-trained DeepLabCut (DLCCT) demonstrated superior performance over pre-trained DLC (DLCPT) and OpenPose (OPPT).
  • The DLC refinement function further enhanced the accuracy of markerless locomotion evaluation.
  • Custom training and refinement are crucial for optimizing DeepLabCut's pose estimation for gait analysis.

Conclusions:

  • DeepLabCut, particularly with custom training and refinement, is a highly effective markerless solution for locomotion analysis.
  • This study provides essential insights into DLC training for optimal performance in movement assessment.
  • Clinicians and practitioners can utilize these findings for accurate, affordable movement analysis beyond laboratory settings.