Research map / Label-efficient and robust learning / Domain adaptation and generalization
SelfReg: Self-Supervised Contrastive Regularization for Domain Generalization
Daehee Kim, Youngjun Yoo, Seunghyun Park, Jinkyu Kim, Jaekoo Lee
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 IDCL (ablation) | SelfReg · PACS | accuracy: 83.30 -> 83.62 | Improved |
| add SWA (ablation) | SelfReg · PACS | accuracy: 83.05 -> 83.30 | Improved |
| add CDPL (ablation) | SelfReg · PACS | accuracy: 82.84 -> 83.05 | Improved |
| add two-domain mixup (ablation) | SelfReg · PACS | accuracy: 82.75 -> 82.84 | Improved |
| add feature-level dissimilarity loss (ablation) | SelfReg · PACS | accuracy: 82.62 -> 82.75 | Improved |
| add logit-level dissimilarity loss (ablation) | baseline ERM · PACS | accuracy: 81.12 -> 82.62 | Improved |
| add SelfReg to ERM | ERM ResNet-18 · PACS | accuracy: 81.12 -> 83.62 | Improved |
| replace RSC with SelfReg | ResNet-18 · PACS | accuracy: 82.10 -> 83.62 | Improved |
| replace RSC with SelfReg | single-source training · PACS single-source | accuracy: 58.66 -> 59.59 | Improved |
| add SelfReg to ERM | ResNet-50 DomainBed · DomainNet | accuracy: 40.9 -> 42.8 | Improved |
| add SelfReg to ERM | ResNet-50 DomainBed · OfficeHome | accuracy: 66.5 -> 67.9 | Improved |
| add SelfReg to ERM | ResNet-50 DomainBed · DomainBed average over 7 datasets | accuracy: 66.6 -> 67.3 | Improved |
Later papers that compared with this method
4 later accepted papers report SelfReg as a baseline in their results tables; 4 of them report a better number on most of the results they share with it.
- Bringing Equity to Classification: Domain Generalization for Domain-Linked Classes (TPAMI 2025 · better on 3 of 3 shared results)
- Causal Balancing for Domain Generalization (ICLR 2023 · better on 1 of 1 shared result)
- PCL: Proxy-Based Contrastive Learning for Domain Generalization (CVPR 2022 · better on 1 of 1 shared result)
- Attention Diversification for Domain Generalization (ECCV 2022 · 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
- DomCLP: Domain-wise Contrastive Learning with Prototype Mixup for Unsupervised Domain Generalization (AAAI 2025 · similarity 0.89)
- PCL: Proxy-Based Contrastive Learning for Domain Generalization (CVPR 2022 · similarity 0.89)
- MetaReg: Towards Domain Generalization using Meta-Regularization (NeurIPS 2018 · similarity 0.88)
- Negatives-Dominant Contrastive Learning for Generalization in Imbalanced Domains (ICML 2026 · similarity 0.88)
- Improving Generalization With Domain Convex Game (CVPR 2023 · similarity 0.88)
- Towards domain-invariant Self-Supervised Learning with Batch Styles Standardization (ICLR 2024 · 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.
