Universal Domain Adaptation 벤치마크
Universal Domain Adaptation on VisDA2017
H-score
- 2019-06-01 — UAN: H-score 34.8
- 2020-02-20 — SHOT-O: H-score 44.0
- 2021-04-07 — OVANet: H-score 53.1
- 2021-12-16 — UMAD: H-score 58.3
- 2023-01-01 — SAN: H-score 60.1
- 2023-03-13 — GLC: H-score 73.1
- 2024-03-06 — LEAD: H-score 76.6
- 2025-06-04 — TASC: H-score 90.36
| Rank | Model | H-score | Source-free | Paper | Code | Year |
|---|---|---|---|---|---|---|
| 21 | UniOT | 57.32 | no | Unified Optimal Transport Framework for Universal Domain Adaptation | changwxx/uniot-for-unida | 2022 |
| 22 | GATE | 56.4 | no | Geometric Anchor Correspondence Mining With Uncertainty Modeling for Universal Domain Adaptation | 2022 | |
| 23 | OVANet | 53.1 | no | OVANet: One-vs-All Network for Universal Domain Adaptation | VisionLearningGroup/OVANet · VisionLearningGroup/visda21-dev | 2021 |
| 24 | SHOT-O | 44.0 | yes | Do We Really Need to Access the Source Data? Source Hypothesis Transfer for Unsupervised Domain Adaptation | tim-learn/SHOT · Claydon-Wang/OFTTA · AlexisGuichemerreCode/survey_hist_wsol_sfda | 2020 |
| 25 | DCC | 43.0 | no | Domain Consensus Clustering for Universal Domain Adaptation | Solacex/Domain-Consensus-Clustering | 2021 |
| 26 | UAN | 34.8 | no | Universal Domain Adaptation | 2019 |