Research map / Image generation / Image restoration
Noise-Modeled Diffusion Models for Low-Light Spike Image Restoration
Ruonan Liu, Lin Zhu, Xijie Xiang, Lizhi Wang, Hua Huang
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 the difference Id to the deterministic residual component of the diffusion process (ablation) | Noise-modeled spike diffusion · Simulated low-light spike dataset | PSNR: 23.32 -> 26.80 | Improved |
| model the diffusion as spike degradation with proportional and residual terms instead of direct residual shift y = x + n (ablation) | Noise-modeled spike diffusion · Simulated low-light spike dataset | PSNR: 23.14 -> 27.16 | Improved |
| decouple the proportional term from the residual alone (y = Ax + n) (ablation) | Noise-modeled spike diffusion · Simulated low-light spike dataset | PSNR: 23.14 -> 23.05 | No clear effect |
| add the proportional term A once deterministic residual with Id is present (ablation) | Noise-modeled spike diffusion · Simulated low-light spike dataset | PSNR: 26.80 -> 27.16 | Improved |
| use inverse shifting schedule (small p for non-deterministic residual) rather than opposite schedule (ablation) | Noise-modeled spike diffusion · Simulated low-light spike dataset | PSNR: 22.38 -> 27.16 | Improved |
| add all three model conditions (region mask, dark current map, spike density value) (ablation) | Noise-modeled spike diffusion · Simulated low-light spike dataset | PSNR: 25.46 -> 27.16 | Improved |
| retrain spike reconstruction baselines on low-light TFP-window data | Spk2ImgNet · Simulated low-light spike dataset | PSNR: 5.94 -> 20.26 | Improved |
| generic diffusion restoration (ResShift) vs retrained regression baselines on simulated data | ResShift · Simulated low-light spike dataset | PSNR: 21.66 vs 21.53 | No clear effect |
| proposed method vs Hu et al. on real data CLIPIQA | Noise-modeled spike diffusion · Real low-light spike dataset | CLIPIQA: 0.3063 vs 0.3003 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- ResShift — ResShift: Efficient Diffusion Model for Image Super-resolution by Residual Shifting (NeurIPS 2023) · better on 4 of 4 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Robust Noise Modeling for Spike Camera via Time-Interval Quantification and Spike-DSLR Multimodal Dataset in Low-Light Imaging (AAAI 2026 · similarity 0.89)
- SpikeReveal: Unlocking Temporal Sequences from Real Blurry Inputs with Spike Streams (NeurIPS 2024 · similarity 0.87)
- Rethinking High-speed Image Reconstruction Framework with Spike Camera (AAAI 2025 · similarity 0.87)
- Joint Demosaicing and Denoising for Spike Camera (AAAI 2024 · similarity 0.86)
- USP-Gaussian: Unifying Spike-based Image Reconstruction, Pose Correction and Gaussian Splatting (CVPR 2025 · similarity 0.86)
- Boosting Spike Camera Image Reconstruction from a Perspective of Dealing with Spike Fluctuations (CVPR 2024 · similarity 0.85)
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.
