Object-Level OOD Detection on STU 1.0 (test)
26.14RecallQMaxLogit
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| MaxLogitAuxiliary OOD Data=No, Architecture=Mask4Former [68]2026.04 | 26.14 | 83.06 | 2.13 | 21.71 | 1.77 | |
| NDP-EEAuxiliary OOD Data=Yes, Architecture=Mask4Former [68]2026.04 | 25.58 | 79.93 | 31.26 | 20.44 | 24.99 | |
| RbAAuxiliary OOD Data=No, Architecture=Mask4Former [68]2026.04 | 24.04 | 83.28 | 3.23 | 20.02 | 2.69 | |
| NDP-EnergyAuxiliary OOD Data=Yes, Architecture=Mask4Former [68]2026.04 | 20.31 | 78.21 | 26.94 | 15.88 | 21.07 | |
| Void ClassifierAuxiliary OOD Data=Yes, Architecture=Mask4Former [68]2026.04 | 17.64 | 84.4 | 8.19 | 14.89 | 6.91 | |
| Deep EnsembleAuxiliary OOD Data=No, Architecture=Mask4Former [68]2026.04 | 16.75 | 84.49 | 10.43 | 14.16 | 8.81 | |
| NDP-EntropyAuxiliary OOD Data=Yes, Architecture=Mask4Former [68]2026.04 | 16.33 | 74.4 | 15.09 | 12.15 | 11.22 | |
| MC DropoutAuxiliary OOD Data=No, Architecture=Mask4Former [68]2026.04 | 2.25 | 86.72 | 1.95 | 2.14 | 1.86 |