比较RAW与RGB域图像修复效果，发现模型与成像流程对齐更关键。
Benchmarking RAW and RGB Restoration in Image Signal Processors
- 在RAW或sRGB域进行盲修复，对比不同部署位置性能。
- 针对噪声、模糊等退化，预处理阶段修复效果优于通用后处理模型。
- 结合ISP特性的训练策略可实现最佳整体性能，适合移动端图像恢复研究者。
现代相机通过图像信号处理器（ISP）将原始传感器数据转换为sRGB图像。本文对固定ISP前后两个位置的盲修复进行了基准测试：(A) 在RAW域的预ISP修复，(B) 在sRGB域的后ISP修复。评估涵盖四类智能手机设备、两个学习型ISP、三种退化情形——噪声、模糊以及联合噪声与模糊，并比较多个代表性RAW与RGB修复模型。结果表明，修复位置本身并不决定性能高低。预ISP的RAW修复策略优于最佳通用RGB修复模型。然而，经过考虑ISP变换过程训练的RGB修复模型整体表现最优。本研究构建的新基准揭示：图像重建性能高度依赖于修复模型与目标成像流水线的匹配程度。因此建议在实验中报告修复放置位置及是否采用ISP感知监督。代码已开源：https://github.com/mv-lab/AISP。
原文摘要 · Abstract (English)
Modern cameras transform RAW sensor measurements into sRGB images through an image signal processor (ISP). We benchmark two placements for blind restoration around a fixed ISP: (A) pre-ISP restoration in the RAW domain and (B) post-ISP restoration in the sRGB domain. The benchmark covers four smartphone device groups, two learned ISPs, three degradation regimes--noise, blur, and joint noise and blur--, and several representative RAW and RGB restoration models. Our results show that placement alone does not determine performance. The RAW restoration strategy outperforms the best generic RGB restoration models. However, RGB restoration models trained considering the ISP transformations, achieve the best overall performance. Our novel benchmark demonstrates that the image reconstruction performance strongly depends on the alignment between the restoration model and the target imaging pipeline. We consequently recommend reporting restoration placement and ISP-aware supervision as key experimental factors. Our code is available at https://github.com/mv-lab/AISP
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。