Image Classification on CIFAR-100 (Calibration Evaluation)
85.2AccuracyBaseline + LS
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| Baseline + LSArchitecture=ResNet 1522026.03 | 85.2 | 0.0266 | 0.6173 | 1.6245 | 1.1622 | |
| Baseline + FocalArchitecture=ResNet 1522026.03 | 85 | 0.0101 | 0.5426 | 6.2433 | 0.8376 | |
| BaselineArchitecture=ResNet 1522026.03 | 84.9 | 0.0622 | 0.6016 | 9.3696 | 0.4576 | |
| Baseline + TSArchitecture=ResNet 1522026.03 | 84.9 | 0.0169 | 0.563 | 5.4784 | 0.8382 | |
| Baseline + SRCArchitecture=ResNet 1522026.03 | 84.8 | 0.013 | 0.5703 | 4.3361 | 1.0157 | |
| Baseline + SRC + TSArchitecture=ResNet 1522026.03 | 84.8 | 0.0142 | 0.5651 | 4.9407 | 0.8578 |