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: Nov 29, 2025

Author Spotlight: Impact of Intergenic Interactions on Disease-Identifying Dark Biomarkers
03:37

Author Spotlight: Impact of Intergenic Interactions on Disease-Identifying Dark Biomarkers

Published on: March 1, 2024

1.1K

FMixCutMatch for semi-supervised deep learning.

Xiang Wei1, Xiaotao Wei1, Xiangyuan Kong1

  • 1School of Software Engineering Beijing Jiaotong University Beijing, China.

Neural Networks : the Official Journal of the International Neural Network Society
|November 20, 2020
PubMed
Summary
This summary is machine-generated.

Related Concept Videos

Extraction: Advanced Methods00:56

Extraction: Advanced Methods

864
Metal ions can be separated from one another by complexation with organic ligands–the chelating agent– to form uncharged chelates. Here, the chelating agent must contain hydrophobic groups and behave as a weak acid, losing a proton to bind with the metal. Since most organic ligands used in this process are insoluble or undergo oxidation in the aqueous phase, the chelating agent is initially added to the organic phase and extracted into the aqueous phase. The metal-ligand complex is...
864

You might also read

Related Articles

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

Sort by
Same author

Xinfeng Capsule combined with conventional Western medicine for rheumatoid arthritis: Efficacy, safety, and subgroup analysis in a multicenter real-world study.

Medicine·2026
Same author

Evaluation of serine protein kinase HPrK as an antibacterial target in gram-positive bacteria and mycoplasmas.

Microbiology spectrum·2026
Same author

Revolutionizing Wilson disease prognosis: a machine learning approach to predict acute-on-chronic liver failure.

Journal of translational medicine·2025
Same author

Genomic Insights into the Pathogenicity and Drug-Resistance of a <i>Bacillus cereus</i> Isolated from Human Teeth.

Infection and drug resistance·2024
Same author

Use of Ultrasound-Guided Platelet-Rich Plasma in the Treatment of Carpometacarpal Osteoarthritis.

HCA healthcare journal of medicine·2023
Same author

Up-regulation of death receptor 4 and 5 by celastrol enhances the anti-cancer activity of TRAIL/Apo-2L.

Cancer letters·2010

This study introduces FMixCutMatch (FMCmatch), a novel semi-supervised learning (SSL) method that enhances data augmentation using Fourier transforms. FMCmatch achieves state-of-the-art results on multiple benchmarks by combining Fourier mixing and cutting strategies with advanced regularization techniques.

Area of Science:

  • Computer Science
  • Machine Learning
  • Artificial Intelligence

Background:

  • Semi-supervised learning (SSL) leverages limited labeled data with abundant unlabeled data for model training.
  • Mixed Sample Augmentation (MSA) techniques, like Cut-Mix, have shown efficacy in smoothing the training space for SSL.
  • Fourier space-based data augmentation offers a novel approach to enhance sample diversity and model robustness.

Purpose of the Study:

  • To propose FMixCutMatch (FMCmatch), an advanced SSL method incorporating Fourier space-based data mixing (FMix) and cutting (FCut).
  • To improve labeled and unlabeled data augmentation strategies within the SSL framework.
  • To enhance training efficiency and model convergence through novel regularization and data mixing techniques.

Main Methods:

Keywords:
Mixed sample augmentationRegularizationSemi-supervised learningSoft pseudo-labels

Related Experiment Videos

Last Updated: Nov 29, 2025

Author Spotlight: Impact of Intergenic Interactions on Disease-Identifying Dark Biomarkers
03:37

Author Spotlight: Impact of Intergenic Interactions on Disease-Identifying Dark Biomarkers

Published on: March 1, 2024

1.1K
  • FMixCutMatch combines FMix and FCut for data augmentation, utilizing Fourier transforms in both mixing and cutting processes.
  • The approach generates soft pseudo-labels for unlabeled data and trains the model for consistency with these labels.
  • It employs weighted cross-entropy minimization with masked sample pairs from FMix and introduces batch label distribution entropy maximization and sample confidence entropy minimization for regularization.
  • A dynamic labeled-unlabeled data mixing (DDM) strategy is utilized to accelerate model convergence.

    • Main Results:

    • FMCmatch achieves state-of-the-art performance across various SSL benchmarks including CIFAR-10/100, SVHN, and Mini-Imagenet.
    • The method demonstrates superior results with different network architectures like CNN-13, WRN-28-2, and ResNet-18.
    • Specific achievements include a 4.54% test error on CIFAR-10 (4K labels) and a 41.25% Top-1 test error on Mini-Imagenet (10K labels).

    Conclusions:

    • FMixCutMatch represents a significant advancement in semi-supervised learning, offering robust data augmentation and improved training efficiency.
    • The integration of Fourier space augmentation techniques with advanced regularization and mixing strategies leads to state-of-the-art performance.
    • The proposed method effectively addresses challenges in SSL by maximizing the utility of both labeled and unlabeled data.