Research map / Image generation / GANs
A high-quality robust diffusion framework for corrupted dataset
Quan Dao, Binh Ta, Tung Pham, Anh Tran
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 |
|---|---|---|---|
| replace GAN generator with UOT-based generative model in DDGAN | DDGAN · CIFAR-10 + 10% MNIST | FID: 14.77 -> 6.98 | Improved |
| replace GAN generator with UOT-based generative model in DDGAN | DDGAN · CIFAR-10 (clean) | FID: 3.75 -> 2.95 | Improved |
| replace GAN generator with UOT-based generative model in DDGAN | DDGAN · STL-10 | FID: 21.79 -> 11.50 | Improved |
| model q(x_0|x_t) with UOT instead of q(x_{t-1}|x_t) (ablation) | RDUOT · CIFAR-10 + 5% MNIST | FID: 6.94 -> 4.37 | Improved |
| feed x_{t-1} to the potential network instead of x_0 (ablation) | RDUOT · CIFAR-10 + 5% MNIST | FID: 5.93 -> 4.37 | Improved |
| use softplus conjugate (Lipschitz) instead of chi^2 (ablation) | RDUOT · CIFAR-10 + 5% MNIST | FID: 5.04 -> 4.37 | Improved |
| use softplus conjugate instead of chi^2 (ablation) | RDUOT · CIFAR-10 (clean) | FID: 3.93 -> 2.95 | Improved |
| use KL divergence instead of softplus conjugate (ablation) | RDUOT · CIFAR-10 (clean) | FID: best FID 10.11 before loss explodes | Hurt |
| use RDUOT instead of UOTM on flipped-face outliers | UOTM · CE+FCE | FID: 6.72 -> 5.99 | Improved |
| use RDUOT instead of RobustGAN | RobustGAN · CIFAR-10 + 3% MNIST | FID: 10.63 -> 3.43 | Improved |
Closest papers by meaning
- Robust Classification via a Single Diffusion Model (ICML 2024 · similarity 0.90)
- Diffusion Models are Certifiably Robust Classifiers (NeurIPS 2024 · similarity 0.90)
- Improved Diffusion-based Generative Model with Better Adversarial Robustness (ICLR 2025 · similarity 0.89)
- Towards Understanding the Robustness of Diffusion-Based Purification: A Stochastic Perspective (ICLR 2025 · similarity 0.88)
- CAT: Contrastive Adversarial Training for Evaluating the Robustness of Protective Perturbations in Latent Diffusion Models (ICML 2025 · similarity 0.88)
- Bridging Data Gaps in Diffusion Models with Adversarial Noise-Based Transfer Learning (ICML 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.
