Research map / Image generation / Image restoration
Deep High Dynamic Range Imaging with Large Foreground Motions
Shangzhe Wu, Jiarui Xu, Yu-Wing Tai, Chi-Keung Tang
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 ResNet merger instead of Unet | Ours · Kalantari test set | PSNR-T: 40.81 -> 41.65 | Improved |
| use ResNet merger instead of Unet | Ours · Kalantari test set | HDR-VDP-2: 64.88 -> 64.90 | Improved |
| use ResNet merger instead of Unet | Ours · Kalantari test set | SSIM-L: 0.9837 -> 0.9858 | Improved |
| ResNet instead of Unet | Ours · CPU running time | time (s): 11.9 -> 14.7 | Hurt |
| replace optical-flow-aligned CNN merger with flow-free network | Kalantari · Kalantari test set | HDR-VDP-2: 63.98 -> 64.90 | Improved |
| replace flow-based Kalantari with flow-free ResNet | Kalantari · Kalantari test set | PSNR-T: 42.70 -> 41.65 | Hurt |
| replace flow-based Kalantari with flow-free ResNet | Kalantari · CPU running time | time (s): 72.1 -> 14.7 | Improved |
| add homography background alignment before the network (ablation) | Unet/ResNet · qualitative, large background misalignment | visual quality | Improved |
| add more input LDRs (3 to 5) | Ours · Sen's dataset | visual detail: 3 -> 5 | Improved |
| add extra exposure shots to the single medium exposure | Ours · hallucination examples | visual detail | Improved |
Closest papers by meaning
- HDRFlow: Real-Time HDR Video Reconstruction with Large Motions (CVPR 2024 · similarity 0.90)
- Towards Real-World HDR Video Reconstruction: A Large-Scale Benchmark Dataset and A Two-Stage Alignment Network (CVPR 2024 · similarity 0.89)
- High Dynamic Range Novel View Synthesis with Single Exposure (ICML 2025 · similarity 0.88)
- HDR-GS: Efficient High Dynamic Range Novel View Synthesis at 1000x Speed via Gaussian Splatting (NeurIPS 2024 · similarity 0.88)
- GaussHDR: High Dynamic Range Gaussian Splatting via Learning Unified 3D and 2D Local Tone Mapping (CVPR 2025 · similarity 0.88)
- Self-Supervised High Dynamic Range Imaging with Multi-Exposure Images in Dynamic Scenes (ICLR 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.
