Image Classification on ImageNet-1k (val) (DsCo Accuracy and Full Accuracy)
70.1DsCo AccuracyDsCo
Evaluation results
| Method | Links | ||
|---|---|---|---|
| DsCoModel=ResNet-101, IPC=700, Labeling=ResNet-18 soft labels2026.03 | 70.1 | — | |
| DsCoModel=ResNet-34, IPC=700, Labeling=ResNet-18 soft labels2026.03 | 69.1 | — | |
| DsCoModel=ResNet-18, IPC=700, Labeling=ResNet-18 soft labels2026.03 | 69 | — | |
| DsCoModel=DenseNet-121, IPC=700, Labeling=ResNet-18 soft labels2026.03 | 67.2 | — | |
| DsCoModel=EfficientNet-B0, IPC=700, Labeling=ResNet-18 soft labels2026.03 | 66.4 | — | |
| Full Dataset TrainingModel=ResNet-18, Labeling=ResNet-18 soft labels2026.03 | — | 69 | |
| Full Dataset TrainingModel=ResNet-34, Labeling=ResNet-18 soft labels2026.03 | — | 69.1 | |
| Full Dataset TrainingModel=ResNet-101, Labeling=ResNet-18 soft labels2026.03 | — | 70.1 | |
| Full Dataset TrainingModel=EfficientNet-B0, Labeling=ResNet-18 soft labels2026.03 | — | 66.8 | |
| Full Dataset TrainingModel=DenseNet-121, Labeling=ResNet-18 soft labels2026.03 | — | 67.6 |