Learning to Defer on ImageNet Overlapped Dog Expert
41.23Error RatePiCCE-CE
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| PiCCE-CENumber of Experts=42026.02 | 41.23 | 89.16 | |
| CELoss Formulation=Vanilla, Number of Experts=42026.02 | 41.56 | 88.37 | |
| PiCCE-CENumber of Experts=82026.02 | 41.72 | 89.23 | |
| PiCCE-CENumber of Experts=122026.02 | 41.77 | 89.14 | |
| PiCCE-CENumber of Experts=162026.02 | 41.79 | 88.94 | |
| PiCCE-OvANumber of Experts=42026.02 | 42.03 | 88.3 | |
| PiCCE-OvANumber of Experts=162026.02 | 42.03 | 88.7 | |
| PiCCE-CENumber of Experts=202026.02 | 42.07 | 88.8 | |
| PiCCE-OvANumber of Experts=202026.02 | 42.1 | 88.77 | |
| PiCCE-OvANumber of Experts=82026.02 | 42.15 | 88.51 | |
| PiCCE-OvANumber of Experts=122026.02 | 42.2 | 88.74 | |
| CELoss Formulation=Vanilla, Number of Experts=82026.02 | 42.65 | 87.99 | |
| CELoss Formulation=Vanilla, Number of Experts=122026.02 | 44.37 | 87.85 | |
| OvALoss Formulation=Vanilla, Number of Experts=42026.02 | 44.56 | 87.3 | |
| CELoss Formulation=Vanilla, Number of Experts=162026.02 | 45.17 | 87.54 | |
| CELoss Formulation=Vanilla, Number of Experts=202026.02 | 46.37 | 87.34 | |
| OvALoss Formulation=Vanilla, Number of Experts=82026.02 | 52.05 | 85.56 | |
| OvALoss Formulation=Vanilla, Number of Experts=122026.02 | 56.64 | 84.51 | |
| OvALoss Formulation=Vanilla, Number of Experts=162026.02 | 59.43 | 83.68 | |
| OvALoss Formulation=Vanilla, Number of Experts=202026.02 | 62.91 | 82.93 |