Image Classification on ImageWoof full-sized (test)
86.4Top-1 AccuracyFull
Evaluation Results
| Method | Links | |
|---|---|---|
| FullIPC (Ratio)=10 (0.8%), Test Model=ConvNet-6, Resolution=256x2562023.11 | 86.4 | |
| Minimax DiffusionIPC (Ratio)=100 (7.7%), Test Model=ResNet-18, Resolution=256x2562023.11 | 65.7 | |
| Minimax DiffusionIPC (Ratio)=100 (7.7%), Test Model=ResNetAP-10, Resolution=256x2562023.11 | 64.5 | |
| Minimax DiffusionIPC (Ratio)=100 (7.7%), Test Model=ConvNet-6, Resolution=256x2562023.11 | 61.1 | |
| Minimax DiffusionIPC (Ratio)=70 (5.4%), Test Model=ResNet-18, Resolution=256x2562023.11 | 58.8 | |
| Minimax DiffusionIPC (Ratio)=70 (5.4%), Test Model=ResNetAP-10, Resolution=256x2562023.11 | 58.3 | |
| Minimax DiffusionIPC (Ratio)=50 (3.8%), Test Model=ResNet-18, Resolution=256x2562023.11 | 57.1 | |
| Minimax DiffusionIPC (Ratio)=50 (3.8%), Test Model=ResNetAP-10, Resolution=256x2562023.11 | 56.3 | |
| Minimax DiffusionIPC (Ratio)=70 (5.4%), Test Model=ConvNet-6, Resolution=256x2562023.11 | 55.7 | |
| Minimax DiffusionIPC (Ratio)=50 (3.8%), Test Model=ConvNet-6, Resolution=256x2562023.11 | 53.9 | |
| Minimax DiffusionIPC (Ratio)=20 (1.6%), Test Model=ResNetAP-10, Resolution=256x2562023.11 | 45.8 | |
| Minimax DiffusionIPC (Ratio)=20 (1.6%), Test Model=ResNet-18, Resolution=256x2562023.11 | 42.5 | |
| Minimax DiffusionIPC (Ratio)=10 (0.8%), Test Model=ResNetAP-10, Resolution=256x2562023.11 | 39.2 | |
| Minimax DiffusionIPC (Ratio)=10 (0.8%), Test Model=ResNet-18, Resolution=256x2562023.11 | 37.6 | |
| Minimax DiffusionIPC (Ratio)=20 (1.6%), Test Model=ConvNet-6, Resolution=256x2562023.11 | 37.6 | |
| Minimax DiffusionIPC (Ratio)=10 (0.8%), Test Model=ConvNet-6, Resolution=256x2562023.11 | 37 | |
| DiTIPC (Ratio)=10 (0.8%), Test Model=ConvNet-6, Resolution=256x2562023.11 | 34.2 | |
| GLaDIPC (Ratio)=10 (0.8%), Test Model=ConvNet-6, Resolution=256x2562023.11 | 33.8 | |
| IDC-1IPC (Ratio)=10 (0.8%), Test Model=ConvNet-6, Resolution=256x2562023.11 | 33.3 | |
| DMIPC (Ratio)=10 (0.8%), Test Model=ConvNet-6, Resolution=256x2562023.11 | 26.9 | |
| HerdingIPC (Ratio)=10 (0.8%), Test Model=ConvNet-6, Resolution=256x2562023.11 | 26.7 | |
| RandomIPC (Ratio)=10 (0.8%), Test Model=ConvNet-6, Resolution=256x2562023.11 | 24.3 | |
| K-CenterIPC (Ratio)=10 (0.8%), Test Model=ConvNet-6, Resolution=256x2562023.11 | 19.4 |