Research map / Vision-language and multimodal / Cross-modal learning
AFUNet: Cross-Iterative Alignment-Fusion Synergy for HDR Reconstruction via Deep Unfolding Paradigm
Xinyue Li, Zhangkai Ni, Wenhan Yang
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 spatial alignment module (SAM) to SFM baseline (ablation) | AFUNet · Kalantari HDR | PSNR-mu: 43.94 -> 44.48 | Improved |
| add cross-fusion module (CFM) to SFM baseline (ablation) | AFUNet · Kalantari HDR | PSNR-mu: 43.94 -> 44.62 | Improved |
| add DCM to SFM baseline (ablation) | AFUNet · Kalantari HDR | PSNR-mu: 43.94 -> 44.45 | Improved |
| combine SAM, CFM and DCM together (ablation) | AFUNet · Kalantari HDR | PSNR-mu: 44.62 -> 44.91 | Improved |
| use alignment-then-fusion order instead of fusion-then-alignment (ablation) | AFUNet · Kalantari HDR | PSNR-mu: 44.72 -> 44.91 | Improved |
| increase unfolding stages from 2 to 4 (ablation) | AFUNet · Kalantari HDR | PSNR-mu: 44.40 -> 44.91 | Improved |
| increase unfolding stages from 4 to 6 (ablation) | AFUNet · Kalantari HDR | PSNR-mu: 44.91 -> 44.93 | No clear effect |
| increase unfolding stages from 4 to 5 (ablation) | AFUNet · Kalantari HDR | PSNR-mu: 44.91 -> 44.85 PSNR-mu; 42.59 -> 42.91 PSNR-l | Mixed |
| AFUNet vs SAFNet on PSNR-l | AFUNet · Kalantari HDR | PSNR-l: 43.18 vs 42.59 | Hurt |
Closest papers by meaning
- Dual-Output Multi-Exposure HDR Reconstruction via SDR Fusion and Gain Map Inverse Tone Mapping (ECCV 2026 · similarity 0.89)
- A Unified HDR Imaging Method With Pixel and Patch Level (CVPR 2023 · similarity 0.89)
- LRHDR: Learning Representation-enhanced HDR Video Reconstruction (CVPR 2026 · 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 0.87)
- Generating Content for HDR Deghosting from Frequency View (CVPR 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.
