Research map / Label-efficient and robust learning / Domain adaptation and generalization
Graph Adaptive Knowledge Transfer for Unsupervised Domain Adaptation
Zhengming Ding, Sheng Li, Ming Shao, Yun Fu
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 |
|---|---|---|---|
| use soft-label MMD instead of hard-label MMD (ablation) | GAKT · Office+Home 12 tasks | accuracy (%): soft-label version improves 1-2% | Improved |
| add joint feature learning and label propagation over LP baseline | LP · Office+Home Ar→Cl | accuracy (%): 20.36 -> 34.49 | Improved |
| add joint subspace learning and soft labels over label propagation on original features | LP · Office-31+Caltech-256 C→W | accuracy (%): 80.34 -> 95.36 | Improved |
| replace separate label propagation and hard labels (LSC) with joint soft-label adaptation | LSC · Office-31+Caltech-256 C→W | accuracy (%): 91.18 -> 95.36 | Improved |
| replace separate label propagation and hard labels (LSC) with joint soft-label adaptation | LSC · Office+Home Ar→Cl | accuracy (%): 31.81 -> 34.49 | Improved |
| our method versus LSC on Office-31 A→C | LSC · Office-31+Caltech-256 A→C | accuracy (%): 88.81 -> 90.18 | Improved |
| our method versus WDAN on Office-31 A→C | WDAN · Office-31+Caltech-256 A→C | accuracy (%): 92.26 -> 90.18 | Hurt |
| our method versus LSC on Office+Home Rw→Cl | LSC · Office+Home Rw→Cl | accuracy (%): 42.88 -> 44.58 | Improved |
| our method versus DHN on Office+Home Rw→Cl | DHN · Office+Home Rw→Cl | accuracy (%): 45.13 -> 44.58 | Hurt |
| vary lambda in range 1 to 10 (ablation) | GAKT · Office-31+Caltech-256 C→D, W→C, D→A | accuracy (%) | No clear effect |
| increase projection dimension (ablation) | GAKT · Office-31+Caltech-256 C→D, W→C, D→A | accuracy (%) | Improved |
Closest papers by meaning
- Open-Set Graph Domain Adaptation via Separate Domain Alignment (AAAI 2024 · similarity 0.90)
- A Theory of Learning Unified Model via Knowledge Integration from Label Space Varying Domains (CVPR 2025 · similarity 0.88)
- Revisiting, Benchmarking and Understanding Unsupervised Graph Domain Adaptation (NeurIPS 2024 · similarity 0.88)
- Towards Unsupervised Open-Set Graph Domain Adaptation via Dual Reprogramming (NeurIPS 2025 · similarity 0.88)
- A General Representation-Based Approach to Multi-Source Domain Adaptation (ICML 2025 · similarity 0.88)
- Smoothness Really Matters: A Simple Yet Effective Approach for Unsupervised Graph Domain Adaptation (AAAI 2025 · similarity 0.87)
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.
