Multi-Task Learning 벤치마크
Multi-Task Learning on NYUv2
Mean IoU
- 2022-02-02 — Nash-MTL: Mean IoU 40.13
- 2024-03-15 — SwinMTL: Mean IoU 58.14
| Rank | Model | Mean IoU | Paper | Code | Year |
|---|---|---|---|---|---|
| 1 | SwinMTL | 58.14 | SwinMTL: A Shared Architecture for Simultaneous Depth Estimation and Semantic Segmentation from Monocular Camera Images | pardistaghavi/swinmtl | 2024 |
| 2 | Nash-MTL | 40.13 | Multi-Task Learning as a Bargaining Game | torchjd/torchjd · avivnavon/nash-mtl · cranial-xix/famo · +1 | 2022 |
| 3 | SwinMTL | 58.14 | SwinMTL: A Shared Architecture for Simultaneous Depth Estimation and Semantic Segmentation from Monocular Camera Images | pardistaghavi/swinmtl | 2024 |
| 4 | Nash-MTL | 40.13 | Multi-Task Learning as a Bargaining Game | torchjd/torchjd · avivnavon/nash-mtl · cranial-xix/famo · +1 | 2022 |