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

Reinforcement learning for dynamic speed control in connected and autonomous vehicles: A review of applications and challenges.

iScience·2026
See all related articles

Related Experiment Video

Updated: Sep 26, 2025

Measuring Statistical Learning Across Modalities and Domains in School-Aged Children Via an Online Platform and Neuroimaging Techniques
08:05

Measuring Statistical Learning Across Modalities and Domains in School-Aged Children Via an Online Platform and Neuroimaging Techniques

Published on: June 30, 2020

7.7K

Survey on Self-Supervised Learning: Auxiliary Pretext Tasks and Contrastive Learning Methods in Imaging.

Saleh Albelwi1,2

  • 1Faculty of Computing and Information Technology, University of Tabuk, Tabuk 47731, Saudi Arabia.

Entropy (Basel, Switzerland)
|April 23, 2022
PubMed
Summary

Self-supervised learning (SSL) offers an alternative to costly deep learning annotations by using unlabeled data. This review covers top SSL methods, including pretext tasks and contrastive learning, for improved feature representation.

Keywords:
auxiliary pretext taskscontrastive learningcontrastive lossdata augmentationdownstream tasksencoderpretext tasksself-supervised learning (SSL)

More Related Videos

Development of a Gaze-Contingent Display Framework Designed for Perceptual and Oculomotor Research with Simulated Central Vision Loss
07:12

Development of a Gaze-Contingent Display Framework Designed for Perceptual and Oculomotor Research with Simulated Central Vision Loss

Published on: April 11, 2025

579
Defining the Role Of Language in Infants' Object Categorization with Eye-tracking Paradigms
07:31

Defining the Role Of Language in Infants' Object Categorization with Eye-tracking Paradigms

Published on: February 8, 2019

6.7K

Related Experiment Videos

Last Updated: Sep 26, 2025

Measuring Statistical Learning Across Modalities and Domains in School-Aged Children Via an Online Platform and Neuroimaging Techniques
08:05

Measuring Statistical Learning Across Modalities and Domains in School-Aged Children Via an Online Platform and Neuroimaging Techniques

Published on: June 30, 2020

7.7K
Development of a Gaze-Contingent Display Framework Designed for Perceptual and Oculomotor Research with Simulated Central Vision Loss
07:12

Development of a Gaze-Contingent Display Framework Designed for Perceptual and Oculomotor Research with Simulated Central Vision Loss

Published on: April 11, 2025

579
Defining the Role Of Language in Infants' Object Categorization with Eye-tracking Paradigms
07:31

Defining the Role Of Language in Infants' Object Categorization with Eye-tracking Paradigms

Published on: February 8, 2019

6.7K

Area of Science:

  • Computer Science
  • Artificial Intelligence
  • Machine Learning

Background:

  • Deep learning models require extensive, costly data annotations.
  • Self-supervised learning (SSL) leverages unlabeled data to overcome annotation limitations.
  • SSL methods learn feature representations through pretext tasks or contrastive learning.

Purpose of the Study:

  • To provide a comprehensive literature review of prominent self-supervised learning methods.
  • To detail the motivation, pipeline, and terminology of SSL.
  • To examine pretext tasks and contrastive learning techniques within SSL.

Main Methods:

  • Review of auxiliary pretext tasks for representation learning.
  • Analysis of contrastive learning approaches in SSL.
  • Comparison of self-supervised methods with traditional supervised learning.

Main Results:

  • SSL methods effectively learn feature representations from unlabeled data.
  • Pretext tasks and contrastive learning are key techniques in successful SSL.
  • SSL shows comparable or improved performance to supervised methods in various downstream tasks.

Conclusions:

  • Self-supervised learning is a viable and efficient alternative to supervised learning for deep learning.
  • Further research is needed to address ongoing challenges in SSL.
  • SSL holds significant potential for advancing AI applications without manual data labeling.