Research map / Label-efficient and robust learning / Domain adaptation and generalization
Divergence Optimization for Noisy Universal Domain Adaptation
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Use the two-head divergence optimization method instead of the strongest listed baseline | Noisy UniDA · OfficeHome P20 | accuracy (%): 77.32 -> 76.10 | No clear effect |
| Use the two-head divergence optimization method instead of the strongest listed baseline | Noisy UniDA · Office P45 | accuracy (%): 56.32 -> 62.49 | Helped |
| Use the two-head divergence optimization method instead of the strongest listed baseline | Noisy UniDA · VisDA S20 | accuracy (%): 62.97 -> 70.53 | Helped |
| Add small-loss selection (ablation) | Ours · Office 6 tasks, P20 noise | accuracy (%): 86.02 -> 91.22 | Helped |
| Add source divergence to source classification (ablation) | Ours · Office 6 tasks, P20 noise | accuracy (%): 90.14 -> 91.22 | Helped |
| Add target cross-entropy component (ablation) | Ours · Office 6 tasks, P20 noise | accuracy (%): 90.30 -> 91.22 | Helped |
| Add target entropy component (ablation) | Ours · Office 6 tasks, P20 noise | accuracy (%): 90.45 -> 91.22 | Helped |
| Separate target divergences (ablation) | Ours · Office 6 tasks, P20 noise | accuracy (%): 88.53 -> 91.22 | Helped |
| Use mini-max classifier-generator training for alignment (ablation) | Ours · Office 6 tasks, P20 noise | accuracy (%): 87.31 -> 91.22 | Helped |
| Use joint divergence rather than symmetric KL for target separation (ablation) | Ours · Office 6 tasks, P20 noise | accuracy (%): 89.75 -> 91.22 | Helped |
Closest papers by meaning
- Universal Semi-Supervised Domain Adaptation by Mitigating Common-Class Bias (CVPR 2024 · similarity 0.91)
- Gradient-Based Sample Selection for Black-Box Universal Domain Adaptation (AAAI 2025 · similarity 0.90)
- MLNet: Mutual Learning Network with Neighborhood Invariance for Universal Domain Adaptation (AAAI 2024 · similarity 0.89)
- Batch Singular Value Polarization and Weighted Semantic Augmentation for Universal Domain Adaptation (ICML 2024 · similarity 0.89)
- Can We Evaluate Domain Adaptation Models Without Target-Domain Labels? (ICLR 2024 · similarity 0.89)
- Low Category Uncertainty and High Training Potential Instance Learning for Unsupervised Domain Adaptation (AAAI 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.