Research map / Label-efficient and robust learning / Domain adaptation and generalization
Image to Image Translation for 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 I2I Adapt to DenseNet (ablation) | DenseNet digit classifier · SVHN → MNIST | accuracy: 80.1 -> 92.1 | Helped |
| Add I2I Adapt to DenseNet (ablation) | DenseNet segmentation model · GTA5 → Cityscapes | mIoU: 29.0 -> 35.7 | Helped |
| Use DenseNet instead of LeNet (ablation) | I2I Adapt · SVHN → MNIST | accuracy: 80.3 -> 92.1 | Helped |
| Use DenseNet instead of dilated ResNet (ablation) | I2I Adapt · GTA5 → Cityscapes | mIoU: 31.8 -> 35.7 | Helped |
| Remove the feature distribution adversarial loss (ablation) | I2I Adapt · Digit adaptation | accuracy | Hurt |
| Add feature distribution adversarial loss (ablation) | Image-to-image adaptation · Digit adaptation | accuracy | Helped |
| Add translation classification loss (ablation) | I2I Adapt · Digit datasets | accuracy | No clear effect |
| Remove cycle consistency loss (ablation) | I2I Adapt · GTA5 → Cityscapes | mIoU | No clear effect |
| Use I2I Adapt instead of the strongest listed baseline | I2I Adapt with LeNet · USPS → MNIST | accuracy: 90.1 -> 87.2 | Hurt |
| Use I2I Adapt instead of the strongest listed baseline | I2I Adapt · Office dataset tasks | accuracy: 75.3 vs 72.6; 52.1 vs 54.9; 71.1 vs 67.1; 50.1 vs 56.0; 99.6 vs 99.2; 96.5 vs 96.4 | Mixed |
Closest papers by meaning
- Unsupervised Image-to-Image Translation Networks (NeurIPS 2017 · similarity 0.91)
- CrDoCo: Pixel-Level Domain Transfer With Cross-Domain Consistency (CVPR 2019 · similarity 0.91)
- Unsupervised Domain Adaptation with Residual Transfer Networks (NeurIPS 2016 · similarity 0.90)
- Unsupervised Domain Adaptation With Similarity Learning (CVPR 2018 · similarity 0.90)
- Unsupervised Pixel-Level Domain Adaptation With Generative Adversarial Networks (CVPR 2017 · similarity 0.90)
- Enhancing Domain Adaptation through Prompt Gradient Alignment (NeurIPS 2024 · similarity 0.88)
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.