Research map / Label-efficient and robust learning / Domain adaptation and generalization
Attract or Distract: Exploit the Margin of Open Set
Qianyu Feng, Guoliang Kang, Hehe Fan, Yi Yang
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 semantic categorical alignment (SCA) (ablation) | open-set DA model (SCA + SCM) · Digits open-set DA (average of 3 tasks) | OS (%): 81.9 -> 84.3 | Improved |
| add semantic contrastive mapping (SCM) (ablation) | open-set DA model (SCA + SCM) · Digits open-set DA (average of 3 tasks) | OS (%): 80.9 -> 84.3 | Improved |
| add SCA (ablation) | open-set DA model (SCA + SCM) · Digits open-set DA (average of 3 tasks) | UNK (%): 85.6 -> 87.9 | Improved |
| add SCM (ablation) | open-set DA model (SCA + SCM) · Digits open-set DA (average of 3 tasks) | UNK (%): 85.5 -> 87.9 | Improved |
| add SCA (ablation) | open-set DA model (SCA + SCM) · Office-31 open-set DA (AlexNet, average of 6 tasks) | OS (%): 86.5 -> 89.7 | Improved |
| add SCM (ablation) | open-set DA model (SCA + SCM) · Office-31 open-set DA (AlexNet, average of 6 tasks) | OS (%): 87.7 -> 89.7 | Improved |
| add SCA (ablation) | open-set DA model (SCA + SCM) · Office-31 open-set DA (AlexNet, average of 6 tasks) | OS* (%): 87.0 -> 90.7 | Improved |
| add SCM (ablation) | open-set DA model (SCA + SCM) · Office-31 open-set DA (AlexNet, average of 6 tasks) | OS* (%): 88.8 -> 90.7 | Improved |
| use adaptive margin instead of static margin (ablation) | open-set DA model (SCA + SCM) · Office-31 A->D (AlexNet) | OS and OS* | Improved |
| set contrastive re-weighting omega to 0.5 rather than 0 (no re-weighting) (ablation) | open-set DA model (SCA + SCM) · Office-31 A->D (AlexNet) | accuracy | Improved |
| use proposed method instead of OSDA+BP | open-set DA model (SCA + SCM) · Digits open-set DA (average of 3 tasks) | UNK (%): 85.9 -> 87.9 | Improved |
| use proposed method instead of OSDA+BP | open-set DA model (SCA + SCM) · Office-31 open-set DA (AlexNet, average of 6 tasks) | OS (%): 80.4 -> 89.7 | Improved |
Closest papers by meaning
- Open-Set Heterogeneous Domain Adaptation: Theoretical Analysis and Algorithm (AAAI 2025 · similarity 0.87)
- COSDA: Counterfactual-based Susceptibility Risk Framework for Open-Set Domain Adaptation (ICML 2025 · similarity 0.86)
- Unveiling the Unknown: Unleashing the Power of Unknown to Known in Open-Set Source-Free Domain Adaptation (CVPR 2024 · similarity 0.86)
- Distinguish Then Exploit: Source-free Open Set Domain Adaptation via Weight Barcode Estimation and Sparse Label Assignment (CVPR 2025 · similarity 0.86)
- Open-Set Graph Domain Adaptation via Separate Domain Alignment (AAAI 2024 · similarity 0.85)
- Revisiting Source-Free Domain Adaptation: Insights into Representativeness, Generalization, and Variety (CVPR 2025 · similarity 0.85)
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.
