Research map / Label-efficient and robust learning / Domain adaptation and generalization
Distant Supervised Centroid Shift: A Simple and Efficient Approach to Visual Domain Adaptation
Jian Liang, Ran He, Zhenan Sun, Tieniu Tan
Check 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 centroid-shift subspace adaptation instead of the strongest listed baseline | MCS · Office31 | average accuracy: 76.3 -> 76.5 | Improved |
| Use MCS instead of JGSA | VGG-FC6 features on Office-Caltech · Office-Caltech | average accuracy: 83.6 -> 85.8 | Improved |
| Use MCS instead of JGSA | VGG-FC7 features on Office-Caltech · Office-Caltech | average accuracy: 82.3 -> 85.2 | Improved |
| Use MCS instead of CDAN+E | ResNet-50 features on Office-Home · Office-Home | average accuracy: 65.8 -> 66.3 | Improved |
| Apply MCS to RevGrad intermediate features | RevGrad on Office31 · Office31 | average accuracy: 81.8 -> 87.8 | Improved |
| Apply MCS to DAN intermediate features | DAN on Office31 · Office31 | average accuracy: 81.7 -> 87.4 | Improved |
| Apply MCS to GTA intermediate features | GTA on digit recognition · MNIST → USPS | accuracy: 96.5 -> 97.8 | Improved |
| Use MCSD instead of StP | Multi-source adaptation · PIE | average accuracy: 54.2 -> 78.0 | Improved |
| Use MCSD instead of MsDA | Multi-source adaptation · Office31 | average accuracy: 84.5 -> 86.0 | Improved |
| Use MCSD instead of StP | Multi-source adaptation · Office-Caltech | average accuracy: 93.6 -> 94.8 | Improved |
| Vary subspace dimensionality (ablation) | MCS · Office31 and Office-Home adaptation tasks | accuracy | No clear effect |
| Increase regularization λ toward very large values (ablation) | MCS · Office31 and Office-Home adaptation tasks | accuracy | Hurt |
Closest papers by meaning
- Learning an Invariant Hilbert Space for Domain Adaptation (CVPR 2017 · similarity 0.89)
- Visual Domain Adaptation by Consensus-Based Transfer to Intermediate Domain (AAAI 2020 · similarity 0.87)
- Balanced Open Set Domain Adaptation via Centroid Alignment (AAAI 2021 · similarity 0.87)
- Joint Domain Alignment and Discriminative Feature Learning for Unsupervised Deep Domain Adaptation (AAAI 2019 · similarity 0.87)
- PUnDA: Probabilistic Unsupervised Domain Adaptation for Knowledge Transfer Across Visual Categories (ICCV 2017 · similarity 0.87)
- A General Representation-Based Approach to Multi-Source Domain Adaptation (ICML 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.
