Research map / Image generation / Image restoration
Dual-Attention-Guided Network for Ghost-Free High Dynamic Range Imaging
Qingsen Yan, Dong Gong, Qinfeng Shi, Anton van den Hengel, Chunhua Shen, Ian D. Reid, Yanning Zhang
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 l1 loss instead of l2 loss (ablation) | AHDRNet · Kalantari dataset (15 test images) | PSNR-mu: 43.0630 -> 43.6172 | Improved |
| use l1 loss instead of l2 loss (ablation) | AHDRNet · Kalantari dataset (15 test images) | PSNR-M: 31.7921 -> 33.0429 | Improved |
| add attention modules on non-reference features (ablation) | AHDRNet · Kalantari dataset (15 test images) | PSNR / ghosting | Improved |
| add dilated convolution in residual dense blocks (ablation) | AHDRNet · Kalantari dataset (15 test images) | PSNR | Improved |
| add global residual learning with reference features (ablation) | AHDRNet · Kalantari dataset (15 test images) | PSNR | Improved |
| use more residual blocks (Deep-RB-Net) instead of DRDBs (ablation) | RB-Net · Kalantari dataset (15 test images) | ghosting | No clear effect |
| add optical-flow alignment before AHDRNet | AHDRNet · Kalantari dataset (15 test images) | PSNR-mu: 43.6172 -> 43.9764 | Mixed |
| replace flow-based merging with attention-guided merging | Kalantari · Kalantari dataset (15 test images) | HDR-VDP-2: 60.5088 -> 62.3044 | Improved |
Closest papers by meaning
- Attention-Guided Network for Ghost-Free High Dynamic Range Imaging (CVPR 2019 · similarity 0.99)
- Ghost-Free High Dynamic Range Imaging with Context-Aware Transformer (ECCV 2022 · similarity 0.90)
- Deep High Dynamic Range Imaging with Large Foreground Motions (ECCV 2018 · similarity 0.89)
- Generating Content for HDR Deghosting from Frequency View (CVPR 2024 · similarity 0.87)
- Self-Supervised High Dynamic Range Imaging with Multi-Exposure Images in Dynamic Scenes (ICLR 2024 · similarity 0.86)
- High Dynamic Range Novel View Synthesis with Single Exposure (ICML 2025 · 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.
