Guardado en:
Detalles Bibliográficos
Autores principales: Okada, Shuntaro, Doi, Kenji, Yoshihashi, Ryota, Kataoka, Hirokatsu, Tanaka, Tomohiro
Formato: Preprint
Publicado: 2024
Materias:
Acceso en línea:https://arxiv.org/abs/2411.12188
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866910016990085120
author Okada, Shuntaro
Doi, Kenji
Yoshihashi, Ryota
Kataoka, Hirokatsu
Tanaka, Tomohiro
author_facet Okada, Shuntaro
Doi, Kenji
Yoshihashi, Ryota
Kataoka, Hirokatsu
Tanaka, Tomohiro
contents We propose a general framework for optimizing noise schedules in diffusion models, applicable to both training and sampling. Our method enforces a constant rate of change in the probability distribution of diffused data throughout the diffusion process, where the rate of change is quantified using a user-defined discrepancy measure. We introduce three such measures, which can be flexibly selected or combined depending on the domain and model architecture. While our framework is inspired by theoretical insights, we do not aim to provide a complete theoretical justification of how distributional change affects sample quality. Instead, we focus on establishing a general-purpose scheduling framework and validating its empirical effectiveness. Through extensive experiments, we demonstrate that our approach consistently improves the performance of both pixel-space and latent-space diffusion models, across various datasets, samplers, and a wide range of number of function evaluations from 5 to 250. In particular, when applied to both training and sampling schedules, our method achieves a state-of-the-art FID score of 2.03 on LSUN Horse 256$\times$256, without compromising mode coverage.
format Preprint
id arxiv_https___arxiv_org_abs_2411_12188
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Constant Rate Scheduling: A General Framework for Optimizing Diffusion Noise Schedule via Distributional Change
Okada, Shuntaro
Doi, Kenji
Yoshihashi, Ryota
Kataoka, Hirokatsu
Tanaka, Tomohiro
Computer Vision and Pattern Recognition
Machine Learning
We propose a general framework for optimizing noise schedules in diffusion models, applicable to both training and sampling. Our method enforces a constant rate of change in the probability distribution of diffused data throughout the diffusion process, where the rate of change is quantified using a user-defined discrepancy measure. We introduce three such measures, which can be flexibly selected or combined depending on the domain and model architecture. While our framework is inspired by theoretical insights, we do not aim to provide a complete theoretical justification of how distributional change affects sample quality. Instead, we focus on establishing a general-purpose scheduling framework and validating its empirical effectiveness. Through extensive experiments, we demonstrate that our approach consistently improves the performance of both pixel-space and latent-space diffusion models, across various datasets, samplers, and a wide range of number of function evaluations from 5 to 250. In particular, when applied to both training and sampling schedules, our method achieves a state-of-the-art FID score of 2.03 on LSUN Horse 256$\times$256, without compromising mode coverage.
title Constant Rate Scheduling: A General Framework for Optimizing Diffusion Noise Schedule via Distributional Change
topic Computer Vision and Pattern Recognition
Machine Learning
url https://arxiv.org/abs/2411.12188