Research map / Image generation / Image restoration
A Unified HDR Imaging Method With Pixel and Patch Level
Qingsen Yan, Weiye Chen, Song Zhang, Yu Zhu, Jinqiu Sun, 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 HyHDRNet instead of CA-ViT | HDR deghosting · Kalantari test | PSNR-μ: 44.32 -> 44.64 | Improved |
| Use HyHDRNet instead of CA-ViT | HDR deghosting · Hu test | PSNR-L: 51.17 -> 51.91 | Improved |
| Add ghost attention (GA) (ablation) | Baseline HDR network · Kalantari test | PSNR-μ: 43.56 -> 43.99 | Improved |
| Add patch aggregation (PA) (ablation) | Baseline HDR network · Kalantari test | PSNR-μ: 43.56 -> 44.22 | Improved |
| Add gating to combine GA and PA (ablation) | HDR network with GA and PA · Kalantari test | PSNR-μ: 44.22 -> 44.49 | Improved |
| Replace gating with addition (ablation) | HDR network with GA and PA · Kalantari test | PSNR-μ: 44.49 -> 44.28 | Hurt |
| Replace gating with concatenation (ablation) | HDR network with GA and PA · Kalantari test | PSNR-μ: 44.49 -> 44.32 | Hurt |
| Add deformable layer (DL) (ablation) | HDR network with GA, PA, and gating · Kalantari test | PSNR-μ: 44.49 -> 44.60 | Improved |
| Replace Transformer-based fusion with DRDB convolution network (ablation) | HyHDRNet · Kalantari test | PSNR-μ: 44.64 -> 44.06 | Hurt |
| Add perceptual loss (ablation) | Model 7 · Kalantari test | PSNR-μ: 44.60 -> 44.64 | Improved |
| Use patch aggregation instead of patch matching (ablation) | HDR patch-based reconstruction · Kalantari test | PSNR-μ: 43.24 -> 44.64 | Improved |
Closest papers by meaning
- Towards Practical and Efficient High-Resolution HDR Deghosting with CNN (ECCV 2020 · similarity 0.92)
- Alignment-free HDR Deghosting with Semantics Consistent Transformer (ICCV 2023 · similarity 0.92)
- Generating Content for HDR Deghosting from Frequency View (CVPR 2024 · similarity 0.90)
- Self-Supervised High Dynamic Range Imaging with Multi-Exposure Images in Dynamic Scenes (ICLR 2024 · similarity 0.89)
- Towards Real-World HDR Video Reconstruction: A Large-Scale Benchmark Dataset and A Two-Stage Alignment Network (CVPR 2024 · similarity 0.88)
- Single Shot HDR Recovery via a Video Diffusion Prior (NeurIPS 2026 · 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.
