Image Classification on STL-10 1000 labels
1.36Error RateSemi-SST
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Semi-SSTBackbone=ViT-Small, Pre-training=DINO2025.05 | 1.36 | — | |
| Super-SSTBackbone=ViT-Small, Pre-training=DINO2025.05 | 1.45 | — | |
| Dash (CTA)Backbone=Wide ResNet-37-2, Training Labels=10002022.05 | 3.96 | — | |
| JEPAMatchIter.=2^172026.04 | 4.28 | — | |
| DoubleMatchBackbone=Wide ResNet-37-2, Training Labels=1000, Evaluation Protocol=min2022.05 | 4.35 | — | |
| DoubleMatch2023.11 | 4.35 | — | |
| DoubleMatchBackbone=Wide ResNet-37-2, Training Labels=1000, Evaluation Protocol=last 202022.05 | 4.46 | — | |
| RegMixMatch*Iter.=2^202026.04 | 4.66 | — | |
| Boosted SSLlabeled samples=10002020.12 | 4.78 | — | |
| FlatMatchIter.=2^202026.04 | 4.82 | — | |
| CrMatchIter.=2^202026.04 | 4.89 | — | |
| DP-SSLBackbone=Wide ResNet-37-2, Training Labels=10002022.05 | 4.97 | — | |
| FixMatchlabeled samples=1000, augmentation=CTA2020.12 | 5.17 | — | |
| FixMatch (CTA)Backbone=Wide ResNet-37-2, Training Labels=10002022.05 | 5.17 | — | |
| ReMixMatchlabeled samples=10002020.12 | 5.23 | — | |
| ReMixMatchBackbone=Wide ResNet-37-2, Training Labels=10002022.05 | 5.23 | — | |
| SequenceMatchBackbone=Wide-ResNet-37-2, #LABEL=10002023.10 | 5.56 | — | |
| NP-Match2023.11 | 5.59 | — | |
| FreeMatchBackbone=Wide-ResNet-37-2, #LABEL=10002023.10 | 5.63 | — | |
| FreeMatch2023.11 | 5.63 | — | |
| FreeMatchIter.=2^202026.04 | 5.63 | — | |
| SimMatchV22023.11 | 5.65 | — | |
| OTAMatch2023.11 | 5.7 | — | |
| CRMatch2025.05 | 5.71 | — | |
| SoftMatch2023.11 | 5.73 | — | |
| SoftMatchIter.=2^202026.04 | 5.73 | — | |
| SimMatchV22025.05 | 5.74 | — | |
| RefixMatch2023.11 | 5.74 | — | |
| FullMatch2023.11 | 5.74 | — | |
| FlexMatchBackbone=Wide-ResNet-37-2, #LABEL=10002023.10 | 5.77 | — | |
| FlexMatchIter.=2^202026.04 | 5.77 | — | |
| Dash2025.05 | 5.82 | — | |
| SimMatch2025.05 | 6.01 | — | |
| AdaMatch2025.05 | 6.08 | — | |
| ReMixMatchBackbone=WRN-37-2, K (Augmentation anchoring parameter)=42019.11 | 6.18 | — | |
| FixMatch2025.05 | 6.23 | — | |
| FixMatchBackbone=Wide-ResNet-37-2, #LABEL=10002023.10 | 6.25 | — | |
| FixMatch2023.11 | 6.25 | — | |
| FixMatchIter.=2^202026.04 | 6.25 | — | |
| CoMatch2025.05 | 6.3 | — | |
| DASHBackbone=Wide-ResNet-37-2, #LABEL=10002023.10 | 6.39 | — | |
| Dash2023.11 | 6.39 | — | |
| UDA2025.05 | 6.44 | — | |
| FlexMatch2025.05 | 6.53 | — | |
| CRMatch2023.11 | 6.53 | — | |
| UDABackbone=Wide-ResNet-37-2, #LABEL=10002023.10 | 6.64 | — | |
| UDAIter.=2^202026.04 | 6.64 | — | |
| MPLBackbone=Wide-ResNet-37-2, #LABEL=10002023.10 | 6.66 | — | |
| MPL2023.11 | 6.66 | — | |
| ReMixMatchBackbone=Wide-ResNet-37-2, #LABEL=10002023.10 | 6.74 | — | |
| ReMixMatchIter.=2^202026.04 | 6.74 | — | |
| ReMixMatchBackbone=WRN-37-2, K (Augmentation anchoring parameter)=12019.11 | 6.77 | — | |
| UDAlabeled samples=10002020.12 | 7.66 | — | |
| UDABackbone=Wide ResNet-37-2, Training Labels=10002022.05 | 7.66 | — | |
| FixMatchlabeled samples=1000, augmentation=RandAugment2020.12 | 7.98 | — | |
| MixMatchBackbone=WRN-37-22019.11 | 10.18 | — | |
| MixMatchlabeled samples=10002020.12 | 10.41 | — | |
| MixMatchBackbone=Wide ResNet-37-2, Training Labels=10002022.05 | 10.41 | — | |
| ReMixMatch2025.05 | 20.17 | — | |
| Mean Teacherlabeled samples=10002020.12 | 21.43 | — | |
| MixMatchBackbone=Wide-ResNet-37-2, #LABEL=10002023.10 | 21.7 | — | |
| MixMatchIter.=2^202026.04 | 21.7 | — | |
| CC-GAN²2019.05 | 22.2 | — | |
| CC-GAN2019.11 | 22.2 | — | |
| SWWAE2019.05 | 25.7 | — | |
| SWWAE2019.11 | 25.7 | — | |
| PI-Modellabeled samples=10002020.12 | 26.23 | — | |
| VAT2025.05 | 27.92 | — | |
| Pseudo-Labelinglabeled samples=10002020.12 | 27.99 | — | |
| Mean Teacher2025.05 | 30.77 | — | |
| Pseudo-Labeling2025.05 | 31.34 | — | |
| Pseudo LabelBackbone=Wide-ResNet-37-2, #LABEL=10002023.10 | 32.64 | — | |
| PseudoLabelIter.=2^202026.04 | 32.64 | — | |
| Pi ModelBackbone=Wide-ResNet-37-2, #LABEL=10002023.10 | 32.78 | — | |
| Pi Model2023.11 | 32.78 | — | |
| MeanTeacherBackbone=Wide-ResNet-37-2, #LABEL=10002023.10 | 33.9 | — | |
| MeanTeacher2023.11 | 33.9 | — | |
| MeanTeacherIter.=2^202026.04 | 33.9 | — | |
| Pi-Model2025.05 | 35.42 | — | |
| VATBackbone=Wide-ResNet-37-2, #LABEL=10002023.10 | 37.95 | — | |
| MixMatch2025.05 | 40.98 | — | |
| CoMatchBackbone=ResNet-182022.03 | — | 79.8 | |
| CoMatchBackbone=Wide ResNet2022.03 | — | 91.5 | |
| FixMatchBackbone=ResNet-182022.03 | — | 65.38 | |
| FixMatchBackbone=Wide ResNet2022.03 | — | 92.46 | |
| MutexMatchBackbone=ResNet-182022.03 | — | 83.36 | |
| MutexMatchBackbone=Wide ResNet2022.03 | — | 92.52 |