Image Classification on CIFAR-10 4000 labels
1.41Error RateSemi-SST
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Semi-SSTBackbone=ViT-Small, Pre-training=DINO2025.05 | 1.41 | — | |
| Super-SSTBackbone=ViT-Small, Pre-training=DINO2025.05 | 1.61 | — | |
| FlexMatch2025.05 | 3.65 | — | |
| SemCoBackbone=WideResnet-28-2, mu=7, shared_codebase=true2021.04 | 3.8 | — | |
| FreeMatchBackbone=Wide-ResNet-28-2, #LABEL=40002023.10 | 4.1 | — | |
| Dash (CTA)Training Labels=40002022.05 | 4.13 | — | |
| SequenceMatchBackbone=Wide-ResNet-28-2, #LABEL=40002023.10 | 4.15 | — | |
| FixMatch+AKC+ARC#label=40002021.03 | 4.19 | — | |
| FlexMatchlabeled samples=40002021.10 | 4.19 | — | |
| FlexMatchBackbone=Wide-ResNet-28-2, #LABEL=40002023.10 | 4.19 | — | |
| FixMatchlabeled samples=40002021.10 | 4.21 | — | |
| FixMatchBackbone=Wide-ResNet-28-2, #LABEL=40002023.10 | 4.21 | — | |
| DP-SSLTraining Labels=40002022.05 | 4.23 | — | |
| FixMatch#label=40002021.03 | 4.24 | — | |
| AdaMatch2025.05 | 4.24 | — | |
| Dash2025.05 | 4.26 | — | |
| CRMatch2025.05 | 4.27 | — | |
| UDAlabeled samples=40002021.10 | 4.29 | — | |
| UDABackbone=Wide-ResNet-28-2, #LABEL=40002023.10 | 4.29 | — | |
| FixMatch (CTA)Training Labels=40002022.05 | 4.31 | — | |
| DASHBackbone=Wide-ResNet-28-2, #LABEL=40002023.10 | 4.36 | — | |
| FixMatch2025.05 | 4.36 | — | |
| CoMatch2025.05 | 4.36 | — | |
| SimMatch2025.05 | 4.36 | — | |
| SimMatchV22025.05 | 4.41 | — | |
| SemCoBackbone=WideResnet-28-2, mu=3, shared_codebase=true2021.04 | 4.43 | — | |
| FixMatchBackbone=WideResnet-28-2, mu=7, shared_codebase=true2021.04 | 4.49 | — | |
| FixMatchBackbone=WideResnet-28-2, mu=3, shared_codebase=true2021.04 | 4.52 | — | |
| MPLBackbone=Wide-ResNet-28-2, #LABEL=40002023.10 | 4.55 | — | |
| Fully-SupervisedBackbone=Wide-ResNet-28-2, #LABEL=40002023.10 | 4.62 | — | |
| DoubleMatchTraining Labels=4000, Evaluation Protocol=min2022.05 | 4.65 | — | |
| UDA2025.05 | 4.65 | — | |
| ReMixMatchTraining Labels=40002022.05 | 4.72 | — | |
| DoubleMatchTraining Labels=4000, Evaluation Protocol=last 202022.05 | 4.83 | — | |
| ReMixMatchlabeled samples=40002021.10 | 4.84 | — | |
| ReMixMatchBackbone=Wide-ResNet-28-2, #LABEL=40002023.10 | 4.84 | — | |
| UDABackbone=WideResnet-28-22021.04 | 4.88 | — | |
| UDATraining Labels=40002022.05 | 4.88 | — | |
| MixMatch+AKC+ARC#label=40002021.03 | 4.92 | — | |
| MixMatch#label=40002021.03 | 5.52 | — | |
| MixMatchBackbone=WideResnet-28-2, shared_codebase=true2021.04 | 6.24 | — | |
| PLCBBackbone=WideResnet-28-22021.04 | 6.28 | — | |
| MixMatchTraining Labels=40002022.05 | 6.42 | — | |
| Mean teacher#label=40002021.03 | 6.43 | — | |
| AKC+ARC#label=40002021.03 | 6.55 | — | |
| ReMixMatch2025.05 | 6.55 | — | |
| MixMatchlabeled samples=40002021.10 | 6.66 | — | |
| MixMatchBackbone=Wide-ResNet-28-2, #LABEL=40002023.10 | 6.66 | — | |
| AKC#label=40002021.03 | 6.72 | — | |
| Pseudo label#label=40002021.03 | 7.04 | — | |
| ARC#label=40002021.03 | 7.07 | — | |
| Supervised labeled#label=40002021.03 | 7.85 | — | |
| Mean Teacherlabeled samples=40002021.10 | 8.1 | — | |
| MeanTeacherBackbone=Wide-ResNet-28-2, #LABEL=40002023.10 | 8.1 | — | |
| VAT2025.05 | 8.29 | — | |
| Mean teacherBackbone=WideResnet-28-22021.04 | 9.19 | — | |
| VATlabeled samples=40002021.10 | 10.51 | — | |
| VATBackbone=Wide-ResNet-28-2, #LABEL=40002023.10 | 10.51 | — | |
| MixMatch2025.05 | 10.9 | — | |
| Label PropagationBackbone=CNN-132021.04 | 12.69 | — | |
| Π-Modellabeled samples=40002021.10 | 13.13 | — | |
| Pi ModelBackbone=Wide-ResNet-28-2, #LABEL=40002023.10 | 13.13 | — | |
| Pseudo-Labeling2025.05 | 13.63 | — | |
| Pseudo-Labelinglabeled samples=40002021.10 | 15.08 | — | |
| Pseudo LabelBackbone=Wide-ResNet-28-2, #LABEL=40002023.10 | 15.08 | — | |
| Mean Teacher2025.05 | 15.32 | — | |
| Pseudo-labelingBackbone=WideResnet-28-22021.04 | 16.09 | — | |
| Pi-Model2025.05 | 16.1 | — | |
| EnAETBackbone=Wide ResNet2022.03 | — | 94.65 | |
| FixMatch(CTA)2022.12 | — | 95.69 | |
| Meta Pseudo Labelsimplementation=original2022.12 | — | 96.11 | |
| Meta Pseudo Labelsimplementation=re-implementation2022.12 | — | 95.87 | |
| MixMatch2022.12 | — | 93.76 | |
| PAWS-NNArchitecture=WideResNet-28-2, Params=1.5M, Epochs=6002021.04 | — | 96 | |
| Self Meta Pseudo Labels2022.12 | — | 95.91 | |
| SimCLRv2Architecture=ResNet-200 (+SK), Params=95M, Epochs=8002021.04 | — | 96 | |
| SimCLRv2Architecture=ResNet-18 (+SK), Params=12M, Epochs=8002021.04 | — | 92.1 | |
| SimPLE2022.12 | — | 94.95 |