Research map / Label-efficient and robust learning / Domain adaptation and generalization
Co-regularized Alignment for Unsupervised Domain Adaptation
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 co-regularized alignment to VADA | VADA · MNIST→SVHN (instance-normalized) | test accuracy: 73.3 -> 81.7 | Helped |
| Add co-regularized alignment to VADA | VADA · STL→CIFAR | test accuracy: 3% | Helped |
| Add co-regularized alignment to VADA | VADA · SVHN→MNIST | test accuracy: about ∼ 1% for no instance-normalization, and ∼ 4% with instance-normalization | Helped |
| Add co-regularized alignment to VADA | VADA · MNIST→MNIST-M | test accuracy: about ∼ 1% for no instance-normalization, and ∼ 2% with instance-normalization | Helped |
| Add co-regularized alignment to VADA | VADA · DIGITS→SVHN | test accuracy: ∼ 1% | Helped |
| Add the explicit diversity loss (ablation) | Co-DA · MNIST→SVHN (instance-normalized) | test accuracy: 78.5/78.2 -> 81.7/80.9 | Helped |
| Use separate batch-normalization layers with shared network parameters (ablation) | Co-DA · MNIST→SVHN (instance-normalized) | test accuracy: 81.7/80.9 vs 81.4/81.3 | No clear effect |
| Share all parameters, including batch normalization, leaving stochastic layers as the only difference (ablation) | Co-DA · MNIST→SVHN (instance-normalized) | test accuracy: 81.7/80.9 vs 79.9 | Hurt |
| Apply DIRT-T refinement after Co-DA (ablation) | Co-DA · MNIST→SVHN (instance-normalized) | test accuracy: 81.7 -> 88.0 | Helped |
| Apply DIRT-T refinement after Co-DA (ablation) | Co-DA · SVHN→MNIST | test accuracy | No clear effect |
Closest papers by meaning
- Unsupervised Domain Adaptation With Similarity Learning (CVPR 2018 · similarity 0.92)
- Unsupervised Domain Adaptation Using Feature-Whitening and Consensus Loss (CVPR 2019 · similarity 0.92)
- Exploiting Local Feature Patterns for Unsupervised Domain Adaptation (AAAI 2019 · similarity 0.92)
- CASUAL: Conditional Support Alignment for Domain Adaptation with Label Shift (AAAI 2025 · similarity 0.91)
- Can We Evaluate Domain Adaptation Models Without Target-Domain Labels? (ICLR 2024 · similarity 0.90)
- Prompt-Based Distribution Alignment for Unsupervised Domain Adaptation (AAAI 2024 · similarity 0.90)
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.