Noise Estimation for Generative Diffusion Models

04/06/2021
by   Robin San-Roman, et al.
0

Generative diffusion models have emerged as leading models in speech and image generation. However, in order to perform well with a small number of denoising steps, a costly tuning of the set of noise parameters is needed. In this work, we present a simple and versatile learning scheme that can step-by-step adjust those noise parameters, for any given number of steps, while the previous work needs to retune for each number separately. Furthermore, without modifying the weights of the diffusion model, we are able to significantly improve the synthesis results, for a small number of steps. Our approach comes at a negligible computation cost.

READ FULL TEXT

page 6

page 7

page 8

research
10/10/2021

Denoising Diffusion Gamma Models

Generative diffusion processes are an emerging and effective tool for im...
research
10/04/2022

Diffusion Models for Graphs Benefit From Discrete State Spaces

Denoising diffusion probabilistic models and score matching models have ...
research
05/25/2023

Parallel Sampling of Diffusion Models

Diffusion models are powerful generative models but suffer from slow sam...
research
11/11/2020

Diffusion Structures for Architectural Stripe Pattern Generation

We present Diffusion Structures, a family of resilient shell structures ...
research
06/09/2023

Boosting Fast and High-Quality Speech Synthesis with Linear Diffusion

Denoising Diffusion Probabilistic Models have shown extraordinary abilit...
research
05/16/2023

Selective Guidance: Are All the Denoising Steps of Guided Diffusion Important?

This study examines the impact of optimizing the Stable Diffusion (SD) g...
research
02/18/2023

Modelos Generativos basados en Mecanismos de Difusión

Diffusion-based generative models are a design framework that allows gen...

Please sign up or login with your details

Forgot password? Click here to reset