Image Classification on STL10 (test) with Quality Metrics (MSE, PSNR, SSIM)
0.003MSEClean
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| Clean2025.12 | 0.003 | 23.43 | 0.732 | 86.5 | |
| Dropoutdelta / lambda=0.972025.12 | 0.005 | 22.7 | 0.676 | 42.4 | |
| DNPdelta / lambda=0.052025.12 | 0.023 | 17.32 | 0.458 | 81.3 | |
| InvL-DNPdelta / lambda=0.062025.12 | 0.026 | 16.74 | 0.427 | 82.5 | |
| Prunedelta / lambda=0.9952025.12 | 0.028 | 16.05 | 0.361 | 74.4 | |
| GNPdelta / lambda=0.032025.12 | 0.037 | 14.58 | 0.29 | 83.1 | |
| InvL-GNPdelta / lambda=0.072025.12 | 0.04 | 14.1 | 0.22 | 84.8 |