Research map / Label-efficient and robust learning / Domain adaptation and generalization
Unknown-Oriented Learning for Open Set Domain Adaptation
Jie Liu, Xiaoqing Guo, Yixuan Yuan
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 multi-unknown detector (additional unknown weights) (ablation) | UOL baseline · Office-Home Ar→Cp | Unk: 48.1 -> 54.3 | Improved |
| add weight discrepancy constraint to multi-unknown detector (ablation) | UOL · Office-Home Ar→Cp | Unk: 54.3 -> 68.8 | Improved |
| add weight discrepancy constraint to multi-unknown detector (ablation) | UOL · Office-Home Ar→Cp | HOS: 53.5 -> 60.6 | Improved |
| add gradient-graph annotation (graph) (ablation) | UOL · Office-Home Ar→Cp | HOS: 60.6 -> 62.0 | Improved |
| use gradient features for the graphs (ablation) | UOL · Office-Home Ar→Cp | Unk: 69.1 -> 73.1 | Improved |
| add kernel-PCA to gradient features (ablation) | UOL · Office-Home Ar→Cp | HOS: 64.1 -> 65.2 | Improved |
| use UOL instead of JPOT | JPOT · Digital recognition (average) | HOS: 88.9 -> 93.1 | Improved |
| use UOL instead of epsilon-OSD | ε-OSD · Office-Home (average) | HOS: 54.0 -> 74.2 | Improved |
| use UOL instead of ROS | ROS · Office-Home (average) | HOS: 66.2 -> 74.2 | Improved |
| use UOL instead of AoD | AoD · Endo-c2k | Unk: 22.2 -> 30.2 | Improved |
| use UOL instead of OSBP | OSBP · Endo-c2k | Unk: 14.3 -> 30.2 | Improved |
| increase additional weights number N within 10 to 20 (ablation) | UOL · Office-Home Ar→Cp (varying openness) | HOS | No clear effect |
Closest papers by meaning
- Separate to Adapt: Open Set Domain Adaptation via Progressive Separation (CVPR 2019 · similarity 0.92)
- Distinguish Then Exploit: Source-free Open Set Domain Adaptation via Weight Barcode Estimation and Sparse Label Assignment (CVPR 2025 · similarity 0.89)
- Open-Set Heterogeneous Domain Adaptation: Theoretical Analysis and Algorithm (AAAI 2025 · similarity 0.89)
- Unveiling the Unknown: Unleashing the Power of Unknown to Known in Open-Set Source-Free Domain Adaptation (CVPR 2024 · similarity 0.89)
- COSDA: Counterfactual-based Susceptibility Risk Framework for Open-Set Domain Adaptation (ICML 2025 · similarity 0.89)
- Dynamic Target Distribution Estimation for Source-Free Open-Set Domain Adaptation (AAAI 2025 · similarity 0.88)
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.
