Research map / Label-efficient and robust learning / Domain adaptation and generalization
Contrastive Vicinal Space for Unsupervised Domain Adaptation
Jaemin Na, Dongyoon Han, Hyung Jin Chang, Wonjun Hwang
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 EMP-Mixup (ablation) | MSTN baseline · Office-31 (avg) | accuracy: 86.5 -> 90.4 | Improved |
| add contrastive loss on top of EMP-Mixup (ablation) | MSTN + EMP-Mixup · Office-31 (avg) | accuracy: 90.4 -> 91.1 | Improved |
| add label-consensus loss on top of EMP-Mixup and contrastive loss (ablation) | MSTN + EMP-Mixup + contrastive loss · Office-31 (avg) | accuracy: 91.1 -> 91.8 | Improved |
| add contrastive loss (ablation) | MSTN + EMP-Mixup · Office-31 A->D | accuracy: 95.6 -> 97.2 | Improved |
| add contrastive loss (ablation) | MSTN + EMP-Mixup · Office-31 A->W | accuracy: 95.9 -> 97.1 | Improved |
| add label-consensus loss (ablation) | MSTN + EMP-Mixup + contrastive loss · Office-31 W->A | accuracy: 76.4 -> 78.4 | Improved |
| add label-consensus loss (ablation) | MSTN + EMP-Mixup + contrastive loss · Office-31 D->A | accuracy: 76.4 -> 77.5 | Improved |
| add EMP-Mixup (ablation) | DANN baseline · Office-31 (avg) | accuracy: 82.2 -> 89.8 | Improved |
| add contrastive loss (ablation) | DANN + EMP-Mixup · Office-31 (avg) | accuracy: 89.8 -> 90.2 | Improved |
| add label-consensus loss (ablation) | DANN + EMP-Mixup + contrastive loss · Office-31 (avg) | accuracy: 90.2 -> 91.0 | Improved |
| apply CoVi instead of MSTN baseline | MSTN · VisDA-C | accuracy: 65.0 -> 88.5 | Improved |
| apply CoVi instead of MSTN baseline | MSTN · PACS (multi-source) | accuracy: 86.95 -> 93.52 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- FixBi — FixBi: Bridging Domain Spaces for Unsupervised Domain Adaptation (CVPR 2021) · better on 3 of 3 shared results
- T-SVDNet — T-SVDNet: Exploring High-Order Prototypical Correlations for Multi-Source Domain Adaptation (ICCV 2021) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Contrastive Adaptation Network for Unsupervised Domain Adaptation (CVPR 2019 · similarity 0.89)
- Contrastive Adaptation Network for Single- and Multi-Source Domain Adaptation (TPAMI 2020 · similarity 0.88)
- Divide and Contrast: Source-free Domain Adaptation via Adaptive Contrastive Learning (NeurIPS 2022 · similarity 0.88)
- SimProF: A Simple Probabilistic Framework for Unsupervised Domain Adaptation (AAAI 2025 · similarity 0.88)
- Low Category Uncertainty and High Training Potential Instance Learning for Unsupervised Domain Adaptation (AAAI 2024 · similarity 0.87)
- Theoretical Performance Guarantees for Partial Domain Adaptation via Partial Optimal Transport (ICML 2025 · similarity 0.86)
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.
