Research map / Image generation / Fast diffusion sampling
GuidNoise: Single-Pair Guided Diffusion for Generalized Noise Synthesis
Changjin Kim, HyeokJun Lee, YoungJoon Yoo
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 a guidance noisy-clean pair to the diffusion model (ablation) | GuidNoise baseline · SIDD-Validation | KLD: 0.080 -> 0.050 | Improved |
| add a guidance noisy-clean pair (ablation) | GuidNoise baseline · SIDD-Validation | AKLD: 0.150 -> 0.118 | Improved |
| add refine loss on top of guidance (ablation) | GuidNoise · SIDD-Validation | KLD: 0.050 -> 0.014 | Improved |
| add refine loss without guidance (ablation) | GuidNoise baseline · SIDD-Validation | AKLD: 0.150 -> 0.163 | Hurt |
| add refine loss on top of guidance (ablation) | GuidNoise · SIDD-Validation | AKLD: 0.118 -> 0.113 | Improved |
| GuidNoise vs NAFlow noise synthesis | noise synthesis · SIDD-Validation | KLD: 0.031 -> 0.014 | Improved |
| GuidNoise vs NAFlow on SIDD+ | noise synthesis · SIDD+ | KLD: 0.049 -> 0.050 | No clear effect |
| GuidNoise vs NeCA-W on Nam | noise synthesis · Nam | KLD: 0.141 -> 0.153 | Hurt |
| train DnCNN on GuidNoise synthetic data instead of NAFlow | DnCNN · SIDD-Validation | PSNR: 37.00 -> 37.07 | Improved |
| add GuidNoise self-augmentation (1/16 real data) (ablation) | NAFNet-Small · SIDD-Validation 1/16 | PSNR (dB): 35.48 (NAFNet-Medium real) -> 35.83 | Improved |
| add self-augmentation (1/8 real data) (ablation) | NAFNet-Small · SIDD-Validation | PSNR (dB): 36.62 vs 36.65 with 1/2 real data | Improved |
Closest papers by meaning
- EVODiff: Entropy-aware Variance Optimized Diffusion Inference (NeurIPS 2025 · similarity 0.88)
- Realistic Noise Synthesis with Diffusion Models (AAAI 2025 · similarity 0.88)
- NoiseDiffusion: Correcting Noise for Image Interpolation with Diffusion Models beyond Spherical Linear Interpolation (ICLR 2024 · similarity 0.88)
- Denoising Diffusion Bridge Models (ICLR 2024 · similarity 0.87)
- NoiseCtrl: A Sampling-Algorithm-Agnostic Conditional Generation Method for Diffusion Models (CVPR 2025 · similarity 0.87)
- Training-Free Adaptive Diffusion with Bounded Difference Approximation Strategy (NeurIPS 2024 · 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.
