arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2303.09508eess.IVcs.CV
收藏

LDMVFI:基于潜在扩散模型的视频帧插值

LDMVFI: Video Frame Interpolation with Latent Diffusion Models

University of Bristol(布里斯托大学)

Duolikun Danier, Fan Zhang, David Bull

AI总结:

本文提出LDMVFI,首次将潜在扩散模型用于视频帧插值,将其建模为条件生成问题,实验表明其感知质量优于现有最先进方法。

摘要:

现有的视频帧插值(VFI)工作大多采用深度神经网络,通过最小化输出与真实帧之间的L1、L2或深度特征空间距离(如VGG损失)来训练。然而,近期研究表明,这些指标并不能很好地反映感知层面的VFI质量。为了开发面向感知质量的VFI方法,本文提出了基于潜在扩散模型的视频帧插值方法LDMVFI。该方法从生成式视角出发,将VFI问题形式化为条件生成问题。作为首个利用潜在扩散模型解决VFI问题的尝试,我们在现有VFI文献常用的测试集上对该方法进行了严格的基准评测。定量实验和用户研究表明,与当前最先进方法相比,LDMVFI即使在分辨率较高的场景下也能插值出感知质量更优的视频内容。代码已开源,地址为https://github.com/danier97/LDMVFI。

英文摘要:

Existing works on video frame interpolation (VFI) mostly employ deep neural networks that are trained by minimizing the L1, L2, or deep feature space distance (e.g. VGG loss) between their outputs and ground-truth frames. However, recent works have shown that these metrics are poor indicators of perceptual VFI quality. Towards developing perceptually-oriented VFI methods, in this work we propose latent diffusion model-based VFI, LDMVFI. This approaches the VFI problem from a generative perspective by formulating it as a conditional generation problem. As the first effort to address VFI using latent diffusion models, we rigorously benchmark our method on common test sets used in the existing VFI literature. Our quantitative experiments and user study indicate that LDMVFI is able to interpolate video content with favorable perceptual quality compared to the state of the art, even in the high-resolution regime. Our code is available at https://github.com/danier97/LDMVFI.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑