List your company
ResearchFormerly WizWand

Image Classification on SVHN 250 labels

1.88Test Error RateCRMatch
0.95127.220613.4919.7594Dec 20, 2019Nov 15, 2020Oct 13, 2021Sep 9, 2022Aug 7, 2023Jul 3, 2024May 31, 2025
Updated 7mo ago

Evaluation results

MethodLinks
CRMatch2023.111.88
SequenceMatchBackbone=Wide-ResNet-28-2, #LABEL=2502023.101.89
UDA2021.101.92
UDABackbone=Wide-ResNet-28-2, #LABEL=2502023.101.92
FixMatch2025.051.93
FreeMatchBackbone=Wide-ResNet-28-2, #LABEL=2502023.101.97
Dash2025.051.97
Dash2023.111.97
FlexMatch2025.051.98
FixMatch2023.111.99
CoMatch2025.052.01
FixMatch2021.102.02
FixMatchBackbone=Wide-ResNet-28-2, #LABEL=2502023.102.02
DASHBackbone=Wide-ResNet-28-2, #LABEL=2502023.102.04
SimMatch2025.052.13
SoftMatch2023.112.15
DP-SSLTraining Labels=2502022.052.16
CRMatch2025.052.16
MPLBackbone=Wide-ResNet-28-2, #LABEL=2502023.102.29
MPL2023.112.29
Boosted SSLlabeled samples=2502020.122.3
DoubleMatchTraining Labels=250, Evaluation Protocol=min2022.052.37
DoubleMatch2023.112.37
Dash (CTA)Training Labels=2502022.052.38
FixMatchlabeled samples=250, augmentation=RandAugment2020.122.48
SimMatchV22025.052.48
DoubleMatchTraining Labels=250, Evaluation Protocol=last 202022.052.58
FixMatchlabeled samples=250, augmentation=CTA2020.122.64
FixMatch (CTA)Training Labels=2502022.052.64
ReMixMatchlabeled samples=2502020.122.92
ReMixMatchTraining Labels=2502022.052.92
SimMatchV22023.112.92
Semi-SSTBackbone=ViT-Small, Pre-training=DINO2025.053.36
MixMatch2025.053.38
VAT2025.053.44
Mean Teacher2021.103.45
MeanTeacherBackbone=Wide-ResNet-28-2, #LABEL=2502023.103.45
Triple-GAN-V2Classifier=13-layer CNN [32], Data Augmentation=Standard2019.123.48
Mean Teacherlabeled samples=2502020.123.57
MixMatchlabeled samples=2502020.123.98
MixMatchTraining Labels=2502022.053.98
FreeMatch2023.114.09
Super-SSTBackbone=ViT-Small, Pre-training=DINO2025.054.17
SNTGClassifier=13-layer CNN (similar), Data Augmentation=Standard2019.124.29
VAT2021.104.33
VATBackbone=Wide-ResNet-28-2, #LABEL=2502023.104.33
MTClassifier=13-layer CNN (similar), Data Augmentation=Standard2019.124.35
MixMatch2021.104.56
MixMatchBackbone=Wide-ResNet-28-2, #LABEL=2502023.104.56
ReMixMatch2025.054.69
MT (our code base)Classifier=13-layer CNN [32], Data Augmentation=Standard2019.124.89
UDA2025.055.65
UDAlabeled samples=2502020.125.69
UDATraining Labels=2502022.055.69
ReMixMatch2021.106.36
ReMixMatchBackbone=Wide-ResNet-28-2, #LABEL=2502023.106.36
FlexMatch2021.106.59
FlexMatchBackbone=Wide-ResNet-28-2, #LABEL=2502023.106.59
AdaMatch2025.058.74
NP-Match2023.119.51
II modelClassifier=13-layer CNN (similar), Data Augmentation=Standard2019.129.69
Π-Model2021.1013.3
Pi ModelBackbone=Wide-ResNet-28-2, #LABEL=2502023.1013.3
Pseudo-Labeling2025.0513.46
Pi Model2023.1113.55
Pseudo-Labeling2021.1015.59
Pseudo LabelBackbone=Wide-ResNet-28-2, #LABEL=2502023.1015.59
Mean Teacher2025.0516.47
PI-Modellabeled samples=2502020.1218.96
Pseudo-Labelinglabeled samples=2502020.1220.21
Pi-Model2025.0524.17
MeanTeacher2023.1125.1