SOTAVerified

Semi-Supervised Image Classification

Semi-supervised image classification leverages unlabelled data as well as labelled data to increase classification performance.

You may want to read some blog posts to get an overview before reading the papers and checking the leaderboards:

( Image credit: Self-Supervised Semi-Supervised Learning )

Papers

Recently AddedMost HypedMost ActiveNeeds VerificationMost Verified

Showing 126–150 of 167 papers

TitleStatusHype
ViTSGMM: A Robust Semi-Supervised Image Recognition Network Using Sparse LabelsCode0
SimMatch: Semi-supervised Learning with Similarity MatchingCode0
Vision Models Are More Robust And Fair When Pretrained On Uncurated Images Without Supervision—0
10,000+ Times Accelerated Robust Subset Selection (ARSS)—0
LiDAM: Semi-Supervised Learning with Localized Domain Adaptation and Iterative Matching—0
OpenMixup: Open Mixup Toolbox and Benchmark for Visual Representation Learning—0
Semi-supervised Sequence-to-sequence ASR using Unpaired Speech and Text—0
Pseudo-Labeling Curriculum for Unsupervised Domain Adaptation—0
Pseudo-label Learning with Calibrated Confidence Using an Energy-based Model—0
Self-Supervised Wasserstein Pseudo-Labeling for Semi-Supervised Image Classification—0
Energy Models for Better Pseudo-Labels: Improving Semi-Supervised Classification with the 1-Laplacian Graph Energy—0
Beyond ADMM: A Unified Client-variance-reduced Adaptive Federated Learning Framework—0
Regularization With Stochastic Transformations and Perturbations for Deep Semi-Supervised Learning—0
Manifold Graph with Learned Prototypes for Semi-Supervised Image Classification—0
Unsupervised High-level Feature Learning by Ensemble Projection for Semi-supervised Image Classification and Image Clustering—0
Applications and Effect Evaluation of Generative Adversarial Networks in Semi-Supervised Learning—0
Diffusion-Based Representation Learning—0
Adversarial Transformations for Semi-Supervised Learning—0
Towards Discovering the Effectiveness of Moderately Confident Samples for Semi-Supervised Learning—0
Graph Convolutional Networks based on Manifold Learning for Semi-Supervised Image Classification—0
How To Overcome Confirmation Bias in Semi-Supervised Image Classification By Active Learning—0
Semi-MAE: Masked Autoencoders for Semi-supervised Vision Transformers—0
Self Adaptive Threshold Pseudo-labeling and Unreliable Sample Contrastive Loss for Semi-supervised Image Classification—0
Weakly Semi-supervised Whole Slide Image Classification by Two-level Cross Consistency Supervision—0
SelfMatch: Combining Contrastive Self-Supervision and Consistency for Semi-Supervised Learning—0
Show:102550
← PrevPage 6 of 7Next →
All datasetsImageNet - 10% labeled dataImageNet - 1% labeled dataCIFAR-10, 4000 Labelscifar-100, 10000 LabelsCIFAR-10, 250 LabelsSVHN, 1000 labelsSVHN, 250 LabelsCIFAR-10, 40 LabelsCIFAR-100, 400 LabelsCIFAR-10, 1000 LabelsCIFAR-100, 2500 LabelsSTL-10, 1000 Labels

Benchmark Results

#ModelMetricClaimedVerifiedStatus
1SimCLR (ResNet-50 4×)Top 5 Accuracy92.6—Unverified
2Rotation + VAT + Ent. Min.Top 5 Accuracy91.23—Unverified
3SimCLR (ResNet-50 2×)Top 5 Accuracy91.2—Unverified
4Mean Teacher (ResNeXt-152)Top 5 Accuracy90.89—Unverified
5OBoW (ResNet-50)Top 5 Accuracy90.7—Unverified
6R2-D2 (ResNet-18)Top 5 Accuracy90.48—Unverified
7FixMatchTop 5 Accuracy89.13—Unverified
8UDATop 5 Accuracy88.52—Unverified
9SimCLR (ResNet-50)Top 5 Accuracy87.8—Unverified
10DHO (ViT-Large)Top 1 Accuracy85.9—Unverified
#ModelMetricClaimedVerifiedStatus
1DHO (ViT-Large)Top 1 Accuracy84.6—Unverified
2OBoW (ResNet-50)Top 5 Accuracy82.9—Unverified
3DHO (ViT-Base)Top 1 Accuracy81.6—Unverified
4REACT (ViT-Large)Top 1 Accuracy81.6—Unverified
5Semi-SST (ViT-Huge)Top 1 Accuracy80.7—Unverified
6Meta Co-TrainingTop 1 Accuracy80.7—Unverified
7Super-SST (ViT-Huge)Top 1 Accuracy80.3—Unverified
8Semi-ViT (ViT-Huge)Top 1 Accuracy80—Unverified
9Semi-ViT (ViT-Large)Top 1 Accuracy77.3—Unverified
10Super-SST (ViT-Small distilled)Top 1 Accuracy76.9—Unverified
#ModelMetricClaimedVerifiedStatus
1Γ-modelPercentage error20.4—Unverified
2GANPercentage error15.59—Unverified
3Bad GANPercentage error14.41—Unverified
4Triple-GAN-V2 (CNN-13, no aug)Percentage error12.41—Unverified
5Pi ModelPercentage error12.16—Unverified
6SESEMI SSL (ConvNet)Percentage error11.65—Unverified
7VATPercentage error11.36—Unverified
8GLOT-DRPercentage error10.6—Unverified
9VAT+EntMinPercentage error10.55—Unverified
10Triple-GAN-V2 (CNN-13)Percentage error10.01—Unverified
#ModelMetricClaimedVerifiedStatus
1Ⅱ-ModelPercentage error39.19—Unverified
2SESEMI SSL (ConvNet)Percentage error38.7—Unverified
3Temporal ensemblingPercentage error38.65—Unverified
4R2-D2 (CNN-13)Percentage error32.87—Unverified
5Dual Student (480)Percentage error32.77—Unverified
6UPS (CNN-13)Percentage error32—Unverified
7SHOT-VAEPercentage error25.3—Unverified
8LiDAMPercentage error23.22—Unverified
9EnAET (WRN-28-2-Large)Percentage error22.92—Unverified
10FixMatch (RA, WRN-28-8)Percentage error22.6—Unverified
#ModelMetricClaimedVerifiedStatus
1Ⅱ-ModelPercentage error53.12—Unverified
2MixUpPercentage error47.43—Unverified
3MeanTeacherPercentage error47.32—Unverified
4VATPercentage error36.03—Unverified
5LiDAMPercentage error19.17—Unverified
6MixMatchPercentage error11.08—Unverified
7RealMixPercentage error9.79—Unverified
8EnAETPercentage error7.6—Unverified
9ReMixMatchPercentage error6.27—Unverified
10FixMatch+CRPercentage error5.04—Unverified