Image Classification Calibration on ImageNet Sketch
0.93CwECEOracle
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| OracleBackbone=ViT-L2026.05 | 0.93 | — | — | — | |
| OracleBackbone=ResNet1522026.05 | 1.64 | — | — | — | |
| OracleBackbone=DenseNet1612026.05 | 1.76 | — | — | — | |
| ECLBackbone=ViT-L2026.05 | 5.18 | — | 1.26 | — | |
| ECLBackbone=ResNet1522026.05 | 9.86 | — | 0.84 | — | |
| ECLBackbone=DenseNet1612026.05 | 10.7 | — | -0.69 | — | |
| ECLBackbone=ResNet1522026.05 | — | — | 0.87 | 9.52 | |
| ECLBackbone=DenseNet1612026.05 | — | — | -0.69 | 10.1 | |
| ECLBackbone=ViT-L2026.05 | — | — | 1.26 | 4.84 |