Universal Domain Adaptation On Visda2017
Metrics
H-score
Source-free
Results
Performance results of various models on this benchmark
Comparison Table
| Model Name | H-score | Source-free |
|---|---|---|
| ovanet-one-vs-all-network-for-universal | 53.1 | no |
| boosting-novel-category-discovery-over-1 | 60.1 | no |
| do-we-really-need-to-access-the-source-data | 44.0 | yes |
| domain-consensus-clustering-for-universal | 43.0 | no |
| lead-learning-decomposition-for-source-free | 76.6 | yes |
| mlnet-mutual-learning-network-with | 69.9 | no |
| universal-domain-adaptation-via-compressive | 65.18 | no |
| umad-universal-model-adaptation-under-domain | 58.3 | yes |
| geometric-anchor-correspondence-mining-with | 56.4 | no |
| unified-optimal-transport-framework-for | 57.32 | no |
| upcycling-models-under-domain-and-category | 73.1 | yes |
| universal-domain-adaptation | 34.8 | no |
| target-semantics-clustering-via-text | 90.36 | no |