Simple Semi-supervised Knowledge Distillation from Vision-Language Models via $\mathbf{\texttt{D}}$ual-$\mathbf{\texttt{H}}$ead $\mathbf{\texttt{O}}$ptimization

Open paper
Benchmark Model Rank Results
semi-supervised-image-classification-on-imagenet-1-labeled-dataDHO (ViT-Large)#1Top 1 Accuracy: 84.6%
semi-supervised-image-classification-on-imagenet-1-labeled-dataDHO (ViT-Base)#3Top 1 Accuracy: 81.6%
semi-supervised-image-classification-on-imagenet-10-labeled-dataDHO (ViT-Large)#1Top 1 Accuracy: 85.9%
semi-supervised-image-classification-on-imagenet-10-labeled-dataDHO (ViT-Base)#6Top 1 Accuracy: 82.8%

Rank counts only results with a code link.