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

Exploring Self-Supervised Vision Transformers for Gait Recognition in the Wild.

Sensors (Basel, Switzerland)·2023
Same author

Learning Gait Representations with Noisy Multi-Task Learning.

Sensors (Basel, Switzerland)·2022
See all related articles

Related Experiment Video

Updated: Oct 8, 2025

Trajectory Data Analyses for Pedestrian Space-time Activity Study
16:14

Trajectory Data Analyses for Pedestrian Space-time Activity Study

Published on: February 25, 2013

13.7K

WildGait: Learning Gait Representations from Raw Surveillance Streams.

Adrian Cosma1, Ion Emilian Radoi1

  • 1Faculty of Automatic Control and Computer Science, University Politehnica of Bucharest, 060042 Bucharest, Romania.

Sensors (Basel, Switzerland)
|December 28, 2021
PubMed
Summary
This summary is machine-generated.

This study introduces WildGait, a novel self-supervised learning framework for gait recognition in real-world scenarios. It enables accurate person identification using only motion data, even with single, uncooperative walking instances.

Keywords:
gait recognitiongraph neural networkspose estimationself-supervised learning

More Related Videos

Combining Eye-tracking Data with an Analysis of Video Content from Free-viewing a Video of a Walk in an Urban Park Environment
08:25

Combining Eye-tracking Data with an Analysis of Video Content from Free-viewing a Video of a Walk in an Urban Park Environment

Published on: May 7, 2019

9.2K
Fully Automated Leg Tracking in Freely Moving Insects using Feature Learning Leg Segmentation and Tracking FLLIT
08:04

Fully Automated Leg Tracking in Freely Moving Insects using Feature Learning Leg Segmentation and Tracking FLLIT

Published on: April 23, 2020

6.9K

Related Experiment Videos

Last Updated: Oct 8, 2025

Trajectory Data Analyses for Pedestrian Space-time Activity Study
16:14

Trajectory Data Analyses for Pedestrian Space-time Activity Study

Published on: February 25, 2013

13.7K
Combining Eye-tracking Data with an Analysis of Video Content from Free-viewing a Video of a Walk in an Urban Park Environment
08:25

Combining Eye-tracking Data with an Analysis of Video Content from Free-viewing a Video of a Walk in an Urban Park Environment

Published on: May 7, 2019

9.2K
Fully Automated Leg Tracking in Freely Moving Insects using Feature Learning Leg Segmentation and Tracking FLLIT
08:04

Fully Automated Leg Tracking in Freely Moving Insects using Feature Learning Leg Segmentation and Tracking FLLIT

Published on: April 23, 2020

6.9K

Area of Science:

  • Computer Science
  • Biometrics
  • Machine Learning

Background:

  • Gait recognition offers non-invasive person identification but traditional methods require controlled environments.
  • Real-world surveillance presents challenges like multiple individuals and single-pass encounters.
  • Existing methods often rely on cooperative scenarios, limiting practical application.

Purpose of the Study:

  • To develop a gait recognition framework for unconstrained, real-world scenarios.
  • To address privacy concerns by focusing solely on motion information.
  • To improve gait recognition accuracy with limited annotated data.

Main Methods:

  • Proposed WildGait, a self-supervised learning framework using Spatio-Temporal Graph Convolutional Networks.
  • Pre-trained the network on a large dataset of automatically annotated, anonymized 2D skeleton sequences (Uncooperative Wild Gait dataset).
  • Utilized motion information from walking individuals, excluding appearance-based data.

Main Results:

  • Achieved state-of-the-art performance in pose-based gait recognition.
  • Demonstrated reliable gait recognition in unconstrained environments.
  • Successfully learned gait signatures from a large dataset of anonymized walking skeletons.

Conclusions:

  • WildGait offers a robust solution for gait recognition in challenging, real-world surveillance.
  • The Uncooperative Wild Gait dataset advances research in unconstrained biometrics.
  • Self-supervised learning effectively addresses data scarcity in gait recognition applications.