Research map / Image generation / Image restoration
Event-guided HDR Reconstruction with Diffusion Priors
Yixin Yang, Jiawei Zhang, Yang Zhang, Yunxuan Wei, Dongqing Zou, Jimmy S. Ren, Boxin Shi
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 |
|---|---|---|---|
| condition diffusion on pretrained event-image encoder features via pyramid fusion instead of concatenating events and LDR image (ablation) | event-guided diffusion HDR · Synthetic HDR | PSNR: 22.60 -> 23.29 | Improved |
| condition on event-image encoder features instead of the HDRev restored image (ablation) | event-guided diffusion HDR · Synthetic HDR | PSNR: 20.79 -> 23.29 | Improved |
| add pyramid fusion module over multi-level HDR features (ablation) | conditioning module · Synthetic HDR | PSNR: 21.62 -> 23.29 | Improved |
| add detail refinement module with histogram-based structure loss (ablation) | diffusion-only pipeline · Synthetic HDR | PSNR: 23.29 -> 25.67 | Improved |
| supervise refinement with local-histogram-adjusted GT instead of raw GT (ablation) | refinement module · Synthetic HDR | PSNR vs raw GT: 26.65 -> 25.67 PSNR, NIQE 3.90 -> 3.86 | Mixed |
| supervise refinement with raw GT (ablation) | refinement module · qualitative | color artifacts and detail loss | Hurt |
| event-guided diffusion approach vs best prior single-image method on synthetic data | Dille et al. · Synthetic HDR | PSNR: 19.73 -> 25.67 | Improved |
| proposed method on HES-HDR no-reference quality | HES-HDR · HES-HDR | NIQE: 4.81 vs NeurImg 4.61 | No clear effect |
Closest papers by meaning
- Single Shot HDR Recovery via a Video Diffusion Prior (NeurIPS 2026 · similarity 0.92)
- EvHDR-GS: Event-guided HDR Video Reconstruction with 3D Gaussian Splatting (AAAI 2025 · similarity 0.89)
- GMODiff: One-Step Gain Map Refinement with Diffusion Priors for Efficient HDR Reconstruction (ECCV 2026 · similarity 0.89)
- Generating Content for HDR Deghosting from Frequency View (CVPR 2024 · similarity 0.88)
- Color Event Enhanced Single-Exposure HDR Imaging (AAAI 2024 · similarity 0.88)
- Self-Supervised High Dynamic Range Imaging with Multi-Exposure Images in Dynamic Scenes (ICLR 2024 · similarity 0.86)
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.
