Image Classification on CIFAR-10 (test) (Accuracy and Loss)
94.39AccuracyDCQ
Evaluation results
| Method | Links | ||
|---|---|---|---|
| DCQBackbone=ResNet-18, Compression Ratio=80%2026.02 | 94.39 | — | |
| DUALBackbone=ResNet-18, Compression Ratio=80%2026.02 | 91.42 | — | |
| DCQBackbone=ResNet-18, Compression Ratio=87.5%2026.02 | 91.02 | — | |
| ADQBackbone=ResNet-18, Compression Ratio=80%2026.02 | 90.4 | — | |
| DQALBackbone=ResNet-18, Compression Ratio=80%2026.02 | 90.2 | — | |
| DQBackbone=ResNet-18, Compression Ratio=80%2026.02 | 89.4 | — | |
| DCQBackbone=ResNet-18, Compression Ratio=92%2026.02 | 89.15 | — | |
| DUALBackbone=ResNet-18, Compression Ratio=87.5%2026.02 | 88.99 | — | |
| DQBackbone=ResNet-18, Compression Ratio=87.5%2026.02 | 87.7 | — | |
| DQALBackbone=ResNet-18, Compression Ratio=87.5%2026.02 | 87.61 | — | |
| ADQBackbone=ResNet-18, Compression Ratio=87.5%2026.02 | 87.51 | — | |
| ADQBackbone=ResNet-18, Compression Ratio=92%2026.02 | 86.32 | — | |
| DUALBackbone=ResNet-18, Compression Ratio=92%2026.02 | 86.15 | — | |
| DQALBackbone=ResNet-18, Compression Ratio=92%2026.02 | 85.3 | — | |
| DQBackbone=ResNet-18, Compression Ratio=92%2026.02 | 84.25 | — | |
| DCQBackbone=ResNet-18, Compression Ratio=96%2026.02 | 79.9 | — | |
| DQBackbone=ResNet-18, Compression Ratio=96%2026.02 | 77.92 | — | |
| ADQBackbone=ResNet-18, Compression Ratio=96%2026.02 | 75.99 | — | |
| DQALBackbone=ResNet-18, Compression Ratio=96%2026.02 | 75.8 | — | |
| DUALBackbone=ResNet-18, Compression Ratio=96%2026.02 | 74.83 | — | |
| OEHGModel=MLP (Multilayer Perceptron)2026.02 | 39.96 | 1.7297 | |
| OEHGModel=SR (Softmax Regression)2026.02 | 38.88 | 1.7693 | |
| T-RHGModel=MLP (Multilayer Perceptron)2026.02 | 35.6 | 1.8475 | |
| RHGModel=MLP (Multilayer Perceptron)2026.02 | 35.02 | 1.9024 | |
| T-RHGModel=SR (Softmax Regression)2026.02 | 32.59 | 1.9686 | |
| RHGModel=SR (Softmax Regression)2026.02 | 31.76 | 1.9497 |