Network Bending of Diffusion Models for Audio-Visual Generation

Fuente: arXiv
Enregistré dans:
Détails bibliographiques
Auteurs principaux: Dzwonczyk, Luke, Cella, Carmine Emanuele, Ban, David
Format: Preprint
Publié: 2024
Sujets:
Accès en ligne:
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866929402629062656
author Dzwonczyk, Luke
Cella, Carmine Emanuele
Ban, David
author_facet Dzwonczyk, Luke
Cella, Carmine Emanuele
Ban, David
contents In this paper we present the first steps towards the creation of a tool which enables artists to create music visualizations using pre-trained, generative, machine learning models. First, we investigate the application of network bending, the process of applying transforms within the layers of a generative network, to image generation diffusion models by utilizing a range of point-wise, tensor-wise, and morphological operators. We identify a number of visual effects that result from various operators, including some that are not easily recreated with standard image editing tools. We find that this process allows for continuous, fine-grain control of image generation which can be helpful for creative applications. Next, we generate music-reactive videos using Stable Diffusion by passing audio features as parameters to network bending operators. Finally, we comment on certain transforms which radically shift the image and the possibilities of learning more about the latent space of Stable Diffusion based on these transforms.
format Preprint
id arxiv_https___arxiv_org_abs_2406_19589
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Network Bending of Diffusion Models for Audio-Visual Generation
Dzwonczyk, Luke
Cella, Carmine Emanuele
Ban, David
Sound
Machine Learning
Multimedia
Audio and Speech Processing
In this paper we present the first steps towards the creation of a tool which enables artists to create music visualizations using pre-trained, generative, machine learning models. First, we investigate the application of network bending, the process of applying transforms within the layers of a generative network, to image generation diffusion models by utilizing a range of point-wise, tensor-wise, and morphological operators. We identify a number of visual effects that result from various operators, including some that are not easily recreated with standard image editing tools. We find that this process allows for continuous, fine-grain control of image generation which can be helpful for creative applications. Next, we generate music-reactive videos using Stable Diffusion by passing audio features as parameters to network bending operators. Finally, we comment on certain transforms which radically shift the image and the possibilities of learning more about the latent space of Stable Diffusion based on these transforms.
title Network Bending of Diffusion Models for Audio-Visual Generation
topic Sound
Machine Learning
Multimedia
Audio and Speech Processing
url https://arxiv.org/abs/2406.19589