Image Classification on MNIST (test) (Accuracy, Calibration, and Statistical Metrics)
99.42AccuracyCDEC-3
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| CDEC-3Backbone=ResNet-502025.12 | 99.42 | 0.0914 | 0.0986 | 0.0033 | 1.3575 | 1.3608 | |
| IDECBackbone=ResNet-502025.12 | 99.31 | 0.0216 | 0.0054 | 0.0162 | 4,298,500,000,000,000 | 4,298,500,000,000,000 | |
| PostNet-3Backbone=ResNet-502025.12 | 99.04 | 0.0388 | 0.0311 | 0.3767 | 3.1372 | 3.5139 | |
| PostNetBackbone=ResNet-502025.12 | 92.24 | 0.0794 | 0.0102 | 2.3888 | 0 | 2.3888 |