发表机构
University of Würzburg（维尔茨堡大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本研究构建覆盖多设备、ISP与退化场景的基准，对比ISP前后的RAW/RGB复原，发现结合ISP变换训练的RGB模型性能最优，建议报告复原位置与ISP感知监督。
AI 中文摘要
现代相机通过图像信号处理器（ISP）将RAW传感器测量值转换为sRGB图像。我们围绕固定ISP对两种盲复原的部署位置进行基准测试：（A）RAW域的ISP前复原，（B）sRGB域的ISP后复原。该基准测试覆盖4组智能手机设备、2种学习型ISP、3种退化场景——噪声、模糊以及噪声与模糊的组合，还有若干代表性RAW与RGB复原模型。结果显示，部署位置本身不决定性能：RAW复原策略优于最优的通用RGB复原模型；但结合ISP变换训练的RGB复原模型，实现了最佳整体性能。我们的新型基准表明，图像重建性能强烈依赖于复原模型与目标成像流水线的匹配度，因此建议将复原部署位置和ISP感知型监督作为关键实验因素。代码可在此https URL获取。
英文摘要
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
CommentsAccepted BMVC 2026: The 37th British Machine Vision Conference
