Noise Crystallization and Liquid Noise: Zero-shot Video Generation using Image Diffusion Models

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Khan, Muhammad Haaris, Reynaud, Hadrien, Kainz, Bernhard
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914966991273984
author Khan, Muhammad Haaris
Reynaud, Hadrien
Kainz, Bernhard
author_facet Khan, Muhammad Haaris
Reynaud, Hadrien
Kainz, Bernhard
contents Although powerful for image generation, consistent and controllable video is a longstanding problem for diffusion models. Video models require extensive training and computational resources, leading to high costs and large environmental impacts. Moreover, video models currently offer limited control of the output motion. This paper introduces a novel approach to video generation by augmenting image diffusion models to create sequential animation frames while maintaining fine detail. These techniques can be applied to existing image models without training any video parameters (zero-shot) by altering the input noise in a latent diffusion model. Two complementary methods are presented. Noise crystallization ensures consistency but is limited to large movements due to reduced latent embedding sizes. Liquid noise trades consistency for greater flexibility without resolution limitations. The core concepts also allow other applications such as relighting, seamless upscaling, and improved video style transfer. Furthermore, an exploration of the VAE embedding used for latent diffusion models is performed, resulting in interesting theoretical insights such as a method for human-interpretable latent spaces.
format Preprint
id arxiv_https___arxiv_org_abs_2410_05322
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Noise Crystallization and Liquid Noise: Zero-shot Video Generation using Image Diffusion Models
Khan, Muhammad Haaris
Reynaud, Hadrien
Kainz, Bernhard
Computer Vision and Pattern Recognition
Although powerful for image generation, consistent and controllable video is a longstanding problem for diffusion models. Video models require extensive training and computational resources, leading to high costs and large environmental impacts. Moreover, video models currently offer limited control of the output motion. This paper introduces a novel approach to video generation by augmenting image diffusion models to create sequential animation frames while maintaining fine detail. These techniques can be applied to existing image models without training any video parameters (zero-shot) by altering the input noise in a latent diffusion model. Two complementary methods are presented. Noise crystallization ensures consistency but is limited to large movements due to reduced latent embedding sizes. Liquid noise trades consistency for greater flexibility without resolution limitations. The core concepts also allow other applications such as relighting, seamless upscaling, and improved video style transfer. Furthermore, an exploration of the VAE embedding used for latent diffusion models is performed, resulting in interesting theoretical insights such as a method for human-interpretable latent spaces.
title Noise Crystallization and Liquid Noise: Zero-shot Video Generation using Image Diffusion Models
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2410.05322