Preserve Your Own Correlation: A Noise Prior for Video Diffusion Models
This work addresses the problem of efficient and high-quality video synthesis for AI applications, representing an incremental improvement by optimizing noise priors in video diffusion models.
The paper tackles the challenge of generating high-quality, temporally coherent videos by finetuning a pretrained image diffusion model with video data, achieving state-of-the-art zero-shot text-to-video results on UCF-101 and MSR-VTT benchmarks with a 10x smaller model and significantly less computation.
Despite tremendous progress in generating high-quality images using diffusion models, synthesizing a sequence of animated frames that are both photorealistic and temporally coherent is still in its infancy. While off-the-shelf billion-scale datasets for image generation are available, collecting similar video data of the same scale is still challenging. Also, training a video diffusion model is computationally much more expensive than its image counterpart. In this work, we explore finetuning a pretrained image diffusion model with video data as a practical solution for the video synthesis task. We find that naively extending the image noise prior to video noise prior in video diffusion leads to sub-optimal performance. Our carefully designed video noise prior leads to substantially better performance. Extensive experimental validation shows that our model, Preserve Your Own Correlation (PYoCo), attains SOTA zero-shot text-to-video results on the UCF-101 and MSR-VTT benchmarks. It also achieves SOTA video generation quality on the small-scale UCF-101 benchmark with a $10\times$ smaller model using significantly less computation than the prior art.