Research map / Label-efficient and robust learning / Domain adaptation and generalization
Learning to Detect Open Classes for Universal Domain Adaptation
Bo Fu, Zhangjie Cao, Mingsheng Long, Jianmin Wang
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add consistency uncertainty (ablation) | CMU · Office-31 (UniDA, avg 6 tasks) | H-score: 72.23 -> 73.14 | Improved |
| add confidence uncertainty (ablation) | CMU · Office-31 (UniDA, avg 6 tasks) | H-score: 72.52 -> 73.14 | Improved |
| add entropy uncertainty (ablation) | CMU · Office-31 (UniDA, avg 6 tasks) | H-score: 69.78 -> 73.14 | Improved |
| add ensemble calibration of entropy and confidence (ablation) | CMU · Office-31 (UniDA, avg 6 tasks) | H-score: 69.50 -> 73.14 | Improved |
| add ensemble calibration of entropy and confidence (ablation) | CMU · Office-31 (UniDA, avg 6 tasks) | per-class accuracy: 88.93 -> 91.11 | Improved |
| add domain similarity to the transferability (ablation) | CMU · Office-31 (UniDA, avg 6 tasks) | H-score: 73.14 -> 73.15 | No clear effect |
| add domain similarity to the transferability (ablation) | CMU · Office-31 (UniDA, avg 6 tasks) | per-class accuracy: 91.11 -> 91.28 | No clear effect |
| use CMU instead of UAN | CMU · Office-31 (UniDA, avg 6 tasks) | H-score: 63.46 -> 73.14 | Improved |
| use CMU instead of UAN | CMU · DomainNet (UniDA, avg 6 tasks) | H-score: 40.98 -> 48.25 | Improved |
| vary threshold w0 within 0.45 to 0.60 (ablation) | CMU · Office-31 A->D | H-score | No clear effect |
Closest papers by meaning
- Unified Optimal Transport Framework for Universal Domain Adaptation (NeurIPS 2022 · similarity 0.92)
- Exploiting Inter-Sample Affinity for Knowability-Aware Universal Domain Adaptation (IJCV 2023 · similarity 0.92)
- Gradient-Based Sample Selection for Black-Box Universal Domain Adaptation (AAAI 2025 · similarity 0.92)
- Universal Semi-Supervised Domain Adaptation by Mitigating Common-Class Bias (CVPR 2024 · similarity 0.92)
- Tackling Dimensional Collapse toward Comprehensive Universal Domain Adaptation (ICML 2025 · similarity 0.90)
- LEAD: Learning Decomposition for Source-free Universal Domain Adaptation (CVPR 2024 · similarity 0.89)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
