Research map / Image generation / Editing and style transfer
Optimal Transport-Guided Conditional Score-Based Diffusion Model
Xiang Gu, Liwei Yang, Jian Sun, Zongben Xu
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 |
|---|---|---|---|
| Use OTCS rather than KNOT | Unpaired super-resolution · CelebA | FID: 5.95 -> 1.77 | Improved |
| Use OTCS rather than NOT | Unpaired super-resolution · CelebA | SSIM: 0.9157 -> 0.9313 | Improved |
| Use OTCS rather than SCONES | Semi-paired image translation · Animal images | FID: 25.24 -> 13.68 | Improved |
| Use OTCS rather than TCR | Semi-paired image translation · Animal images | Acc (%): 40.44 -> 96.44 | Improved |
| Use OTCS rather than KNOT | Semi-paired image translation · Digits | FID: 3.18 -> 5.12 | Hurt |
| Use OTCS rather than TCR | Semi-paired image translation · Digits | Acc (%): 36.21 -> 67.42 | Improved |
| Add resampling-by-compatibility (RSC) (ablation) | OTCS · Animal images and CelebA | FID and SSIM | Improved |
| Increase the number of paired examples (ablation) | OTCS · Animal images | Acc | Improved |
| Increase the number of paired examples (ablation) | OTCS · Animal images | FID | Improved |
| Change the sampling of the same 3 paired examples (ablation) | OTCS · Animal images | FID and Acc | No clear effect |
| Decrease the initial reverse-SDE time M (ablation) | OTCS · CelebA | SSIM | Improved |
| Use smaller M (ablation) | OTCS · CelebA | FID: below 1.85 for M in [0.2, 0.5] | Mixed |
Closest papers by meaning
- Rethinking Diffusion Posterior Sampling: From Conditional Score Estimator to Maximizing a Posterior (ICLR 2025 · similarity 0.88)
- Multistep Distillation of Diffusion Models via Moment Matching (NeurIPS 2024 · similarity 0.88)
- Theory on Score-Mismatched Diffusion Models and Zero-Shot Conditional Samplers (ICLR 2025 · similarity 0.88)
- REG: Rectified Gradient Guidance for Conditional Diffusion Models (ICML 2025 · similarity 0.87)
- DEFT: Efficient Fine-tuning of Diffusion Models by Learning the Generalised $h$-transform (NeurIPS 2024 · similarity 0.87)
- On Statistical Rates of Conditional Diffusion Transformers: Approximation, Estimation and Minimax Optimality (ICLR 2025 · 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.
