Research map / Label-efficient and robust learning / Domain adaptation and generalization
Adaptive Adversarial Network for Source-Free Domain Adaptation
Haifeng Xia, Handong Zhao, Zhengming Ding
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 the soft-adversarial mechanism with a flexible target classifier (ablation) | A2Net · Office-31 D->A and W->A | accuracy | Improved |
| add contrastive category-wise matching (ablation) | A2Net · Office-31 D->A and W->A | accuracy | Improved |
| add self-supervised rotation (ablation) | A2Net · Office-31 D->A and W->A | accuracy | Improved |
| set rotation trade-off eta to 0.3 (ablation) | A2Net · Office-Home (two tasks) | accuracy | Improved |
| flexible adversarially trained target classifier instead of frozen classifier (A2Net vs SHOT) | A2Net · Office-31 D->A | accuracy: 74.7 -> 76.7 | Improved |
| flexible adversarially trained target classifier instead of frozen classifier (A2Net vs SHOT) | A2Net · Office-31 W->A | accuracy: 74.3 -> 76.1 | Improved |
| A2Net vs SFDA on large domain shift | A2Net · Office-Home Ar->Cl | accuracy: 48.4 -> 58.4 | Improved |
| A2Net vs SHOT overall | A2Net · Office-31 (average) | accuracy: 88.6 -> 90.1 | Improved |
| A2Net vs SHOT overall | A2Net · Office-Home (average) | accuracy: 71.8 -> 72.8 | Improved |
| A2Net vs SHOT overall | A2Net · VisDA | per-class accuracy: 82.9 -> 84.3 | Improved |
Closest papers by meaning
- Bridging Theory and Practice in Source-Free Domain Adaptation via Adversarial Proxy Perturbation (ECCV 2026 · similarity 0.92)
- Adversarial Discriminative Domain Adaptation (CVPR 2017 · similarity 0.90)
- Multi-Source Open-Set Deep Adversarial Domain Adaptation (ECCV 2020 · similarity 0.90)
- Adversarial Multiple Source Domain Adaptation (NeurIPS 2018 · similarity 0.90)
- Blending-Target Domain Adaptation by Adversarial Meta-Adaptation Networks (CVPR 2019 · similarity 0.90)
- Transferable Adversarial Training: A General Approach to Adapting Deep Classifiers (ICML 2019 · similarity 0.90)
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.
