Research map / Label-efficient and robust learning / Domain adaptation and generalization
Mitigating Negative Transfer via Reducing Environmental Disagreement
Hui Sun, Zheng Xie, Haoyuan He, Ming Li
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 explicit disagreement-reduction loss L_tr (ablation) | RED · Office-Home (ResNet50) | accuracy: 75.3 -> 76.3 | Improved |
| add both L_dt and L_tr (ablation) | RED · Office-Home (ResNet50) | accuracy: 66.2 -> 76.3 | Improved |
| add FixMatch self-training to CDAN (ablation) | CDAN · Office-Home (ResNet50) | accuracy: 63.8 -> 73.1 | Improved |
| add RED disentanglement and disagreement reduction to CDAN+FixMatch (ablation) | CDAN+FixMatch · Office-Home (ResNet50) | accuracy: 73.1 -> 76.3 | Improved |
| vary alpha and beta in {0.5, 1.0, 1.5, 2.0} (ablation) | RED · Office-Home (ResNet50) | accuracy | No clear effect |
| replace CALE with RED | CALE · Office-Home (ResNet50) | accuracy: 74.8 -> 76.3 | Improved |
| replace CALE with RED | CALE · DomainNet (ResNet50) | avg accuracy: 28.4 -> 37.1 | Improved |
| replace CALE with RED, source domain rel | CALE · DomainNet rel source | avg accuracy: 29.1 -> 41.8 | Improved |
| replace CALE with RED | CALE · Office-Home (DeiT-base) | accuracy: 81.5 -> 82.3 | Improved |
| replace SSRT-B with RED | SSRT-B · Office-Home (ViT-B/16) | accuracy: 85.4 -> 88.0 | Improved |
| replace DAMP multimodal with RED visual-only | DAMP · Office-Home (ViT-B/16) | accuracy: 87.1 -> 88.0 | Improved |
| replace ICON with RED | ICON · Office-31 (ResNet50) | accuracy: 85.1 -> 85.4 | No clear effect |
Closest papers by meaning
- Transporting Causal Mechanisms for Unsupervised Domain Adaptation (ICCV 2021 · similarity 0.88)
- Learning Unbiased Transferability for Domain Adaptation by Uncertainty Modeling (ECCV 2022 · similarity 0.87)
- Unsupervised Robust Domain Adaptation: Paradigm, Theory and Algorithm (IJCV 2025 · similarity 0.87)
- Partial disentanglement for domain adaptation (ICML 2022 · similarity 0.87)
- On the Transferability and Discriminability of Representation Learning in Unsupervised Domain Adaptation (TPAMI 2025 · similarity 0.87)
- Can We Evaluate Domain Adaptation Models Without Target-Domain Labels? (ICLR 2024 · 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.
