Multi-task Learning on Office-31 (test)
89.06Accuracy (Domain A)STL
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| STLWeighting Strategy=N/A2021.02 | 89.06 | 96.72 | 98.89 | 94.89 | — | |
| MOMLWeighting Strategy=N/A2021.02 | 88.03 | 99.18 | 99.44 | 95.55 | — | |
| DMTLWeighting Strategy=EW2021.02 | 87.35 | 99.18 | 98.89 | 95.14 | — | |
| DMTLWeighting Strategy=DWA2021.02 | 86.67 | 99.18 | 97.22 | 94.36 | — | |
| DMTLWeighting Strategy=UW2021.02 | 86.5 | 97.54 | 97.78 | 93.94 | — | |
| DMTLWeighting Strategy=MGDA2021.02 | 81.88 | 97.54 | 98.89 | 92.77 | — |