Semi Supervised Image Classification On Cifar 8

评估指标

Percentage error

评测结果

各个模型在此基准测试上的表现结果

Paper TitleRepository
FixMatch (CTA)49.95±3.01FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence
FixMatch+CR49.23Contrastive Regularization for Semi-Supervised Learning-
Dash (CTA, WRN-28-8)44.83±1.36Dash: Semi-Supervised Learning with Dynamic Thresholding-
Dash (RA, WRN-28-8)44.76±0.96Dash: Semi-Supervised Learning with Dynamic Thresholding-
ReMixMatch44.28±2.06ReMixMatch: Semi-Supervised Learning with Distribution Alignment and Augmentation Anchoring
DP-SSL43.17±1.29DP-SSL: Towards Robust Semi-supervised Learning with A Few Labeled Samples-
PCL (Fixmatch)42.38±2.52Probabilistic Contrastive Learning for Domain Adaptation
DoubleMatch41.83± 1.22DoubleMatch: Improving Semi-Supervised Learning with Self-Supervision
FixMatch+DM40.25±0.95--
FlexMatch39.94±1.62FlexMatch: Boosting Semi-Supervised Learning with Curriculum Pseudo Labeling
CCSSL(FixMatch)38.81Class-Aware Contrastive Semi-Supervised Learning
NP-Match38.67NP-Match: When Neural Processes meet Semi-Supervised Learning
FreeMatch37.98FreeMatch: Self-adaptive Thresholding for Semi-supervised Learning
SimMatch37.81SimMatch: Semi-supervised Learning with Similarity Matching
PCL (Flexmatch)35.75±0.53Probabilistic Contrastive Learning for Domain Adaptation
ShrinkMatch35.36Shrinking Class Space for Enhanced Certainty in Semi-Supervised Learning
ReMixMatch16.8USB: A Unified Semi-supervised Learning Benchmark for Classification
SemiReward15.62SemiReward: A General Reward Model for Semi-supervised Learning
0 of 18 row(s) selected.
Semi Supervised Image Classification On Cifar 8 | SOTA | HyperAI超神经