Saved in:
Bibliographic Details
Main Authors: Ge, Songwei, Nah, Seungjun, Liu, Guilin, Poon, Tyler, Tao, Andrew, Catanzaro, Bryan, Jacobs, David, Huang, Jia-Bin, Liu, Ming-Yu, Balaji, Yogesh
Format: Preprint
Published: 2023
Subjects:
Online Access:https://arxiv.org/abs/2305.10474
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914728104689664
author Ge, Songwei
Nah, Seungjun
Liu, Guilin
Poon, Tyler
Tao, Andrew
Catanzaro, Bryan
Jacobs, David
Huang, Jia-Bin
Liu, Ming-Yu
Balaji, Yogesh
author_facet Ge, Songwei
Nah, Seungjun
Liu, Guilin
Poon, Tyler
Tao, Andrew
Catanzaro, Bryan
Jacobs, David
Huang, Jia-Bin
Liu, Ming-Yu
Balaji, Yogesh
contents 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.
format Preprint
id arxiv_https___arxiv_org_abs_2305_10474
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Preserve Your Own Correlation: A Noise Prior for Video Diffusion Models
Ge, Songwei
Nah, Seungjun
Liu, Guilin
Poon, Tyler
Tao, Andrew
Catanzaro, Bryan
Jacobs, David
Huang, Jia-Bin
Liu, Ming-Yu
Balaji, Yogesh
Computer Vision and Pattern Recognition
Graphics
Machine Learning
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.
title Preserve Your Own Correlation: A Noise Prior for Video Diffusion Models
topic Computer Vision and Pattern Recognition
Graphics
Machine Learning
url https://arxiv.org/abs/2305.10474