LDMVFI: Video Frame Interpolation with Latent Diffusion Models
This addresses the problem of poor perceptual quality in video frame interpolation for applications like video processing, though it is incremental as it adapts existing generative models to a specific task.
The paper tackles video frame interpolation by proposing LDMVFI, a method using latent diffusion models to improve perceptual quality, achieving favorable results compared to state-of-the-art methods in quantitative experiments and user studies.
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.