Research map / Label-efficient and robust learning / Domain adaptation and generalization
Instance Level Affinity-Based Transfer for Unsupervised 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 |
|---|---|---|---|
| Add ILA-DA to CDAN | CDAN · Digits (3 transfer tasks) | average accuracy (%): 93.1 -> 94.88 | Helped |
| Add ILA-DA to DANN | DANN · Digits (3 transfer tasks) | average accuracy (%): 89.29 -> 93.83 | Helped |
| Add ILA-DA to CDAN | CDAN · Office-31 (6 transfer tasks) | average accuracy (%): 86.6 -> 89.30 | Helped |
| Add ILA-DA to DANN | DANN · Office-31 (6 transfer tasks) | average accuracy (%): 82.2 -> 85.54 | Helped |
| Add ILA-DA to CDAN | CDAN · Birds-31 (6 transfer tasks) | average accuracy (%): 82.18 -> 86.03 | Helped |
| Add ILA-DA to DANN | DANN · Birds-31 (6 transfer tasks) | average accuracy (%): 78.44 -> 83.63 | Helped |
| Replace MSC loss with triplet loss (ablation) | CDAN with ILA-DA · Office-31 (6 transfer tasks) | average accuracy (%): 89.30 -> 87.63 | Hurt |
| Replace kNN pseudo-labeling with classifier predictions (ablation) | ILA-DA with CDAN · Office-31 (6 transfer tasks) | average accuracy (%): 89.30 -> 86.68 | Hurt |
| Set sampling fraction μ to 0.33 instead of 0.67 (ablation) | ILA-DA with CDAN · Office-31 (6 transfer tasks) | average accuracy (%): 89.30 -> 87.91 | Hurt |
| Set sampling fraction μ to 0.50 instead of 0.67 (ablation) | ILA-DA with CDAN · Office-31 (6 transfer tasks) | average accuracy (%): 89.30 -> 88.53 | Hurt |
| Accept all pseudo-labels by setting μ to 1.00 (ablation) | ILA-DA with CDAN · Office-31 (6 transfer tasks) | average accuracy (%): 89.30 -> 87.92 | Hurt |
| Use k=1 instead of k=5 neighbors (ablation) | ILA-DA with CDAN · Office-31 (6 transfer tasks) | average accuracy (%): 89.30 -> 87.46 | Hurt |
Closest papers by meaning
- Open-Set Heterogeneous Domain Adaptation: Theoretical Analysis and Algorithm (AAAI 2025 · similarity 0.89)
- Universal Source-Free Domain Adaptation (CVPR 2020 · similarity 0.89)
- Universal Domain Adaptation through Self Supervision (NeurIPS 2020 · similarity 0.89)
- Gradient-Based Sample Selection for Black-Box Universal Domain Adaptation (AAAI 2025 · similarity 0.89)
- Category-Aware Active Domain Adaptation (ICML 2024 · similarity 0.89)
- A Separation and Alignment Framework for Black-Box 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.