Abstract
Existing open-source film restoration methods show limited performance compared to commercial methods due to training with low-quality synthetic data and employing noisy optical flows. In addition, high-resolution films have not been explored by the open-source this http URL propose HaineiFRDM(Film Restoration Diffusion Model), a film restoration framework, to explore diffusion model's powerful content-understanding ability to help human expert better restore indistinguishable film this http URL, we employ a patch-wise training and testing strategy to make restoring high-resolution films on one 24GB-VRAMR GPU possible and design a position-aware Global Prompt and Frame Fusion this http URL, we introduce a global-local frequency module to reconstruct consistent textures among different patches. Besides, we firstly restore a low-resolution result and use it as global residual to mitigate blocky artifacts caused by patching this http URL, we construct a film restoration dataset that contains restored real-degraded films and realistic synthetic this http URL experimental results conclusively demonstrate the superiority of our model in defect restoration ability over existing open-source methods. Code and the dataset will be released.
Abstract (translated)
现有的开源电影修复方法由于使用低质量的合成数据进行训练和采用噪声光学流，其性能相比商业方法较为有限。此外，高分辨率影片的修复尚未被开源社区充分探索。为此，我们提出了HaineiFRDM（Film Restoration Diffusion Model），这是一种基于扩散模型的电影修复框架，旨在利用扩散模型强大的内容理解能力帮助人类专家更好地恢复难以辨别的电影片段。 具体来说，我们采用了一种基于补丁的方法进行训练和测试，使得在配备24GB VRAM GPU上对高分辨率影片进行修复成为可能，并设计了一个位置感知全局提示及帧融合模块。此外，为了重建不同补丁间的连贯纹理，我们引入了全球-局部频率模块。 进一步地，我们在首次恢复低分辨率结果后将其用作全局残差以减轻由分块引起的方块效应（blocky artifacts）。我们构建了一个包含修复的现实降质影片和逼真的合成数据的电影修复数据集。实验结果显示，我们的模型在缺陷修复能力上显著优于现有的开源方法。代码及数据集将公开发布。
URL
https://arxiv.org/abs/2512.24946
