Image Classification on CIFAR-10 (test) (C10/C10+ Reporting)
1.54C10+ Error RateColornet-40-48
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Colornet-40-48No of Parameters=19.0M2019.02 | 1.54 | 3.14 | |
| Densenet-BC-190-40No of Parameters=25.6M2019.02 | 3.46 | — | |
| Colornet-40-12No of Parameters=1.75M2019.02 | 3.49 | 4.98 | |
| Densenet-BC-250-24No of Parameters=15.3M2019.02 | 3.62 | 5.19 | |
| Wide-Resnet-28No of Parameters=36.5M2019.02 | 4.17 | — | |
| Densenet-BC-100-12No of Parameters=0.8M2019.02 | 4.51 | 5.92 | |
| FractalnetNo of Parameters=38.6M, dropout=true2019.02 | 4.6 | 7.33 | |
| Resnet-1202No of Parameters=10.2M2019.02 | 4.91 | — | |
| FractalnetNo of Parameters=38.6M2019.02 | 5.22 | 10.18 | |
| Resnet-101No of Parameters=1.7M2019.02 | 5.23 | 11.66 | |
| all-CNN2019.02 | 7.25 | 9.08 | |
| Highway Network2019.02 | 7.72 | — | |
| N-in-N2019.02 | 8.81 | 10.41 |