Wind Noise Reduction with a Diffusion-based Stochastic Regeneration Model

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Lemercier, Jean-Marie, Thiemann, Joachim, Koning, Raphael, Gerkmann, Timo
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914634681810944
author Lemercier, Jean-Marie
Thiemann, Joachim
Koning, Raphael
Gerkmann, Timo
author_facet Lemercier, Jean-Marie
Thiemann, Joachim
Koning, Raphael
Gerkmann, Timo
contents In this paper we present a method for single-channel wind noise reduction using our previously proposed diffusion-based stochastic regeneration model combining predictive and generative modelling. We introduce a non-additive speech in noise model to account for the non-linear deformation of the membrane caused by the wind flow and possible clipping. We show that our stochastic regeneration model outperforms other neural-network-based wind noise reduction methods as well as purely predictive and generative models, on a dataset using simulated and real-recorded wind noise. We further show that the proposed method generalizes well by testing on an unseen dataset with real-recorded wind noise. Audio samples, data generation scripts and code for the proposed methods can be found online (https://uhh.de/inf-sp-storm-wind).
format Preprint
id arxiv_https___arxiv_org_abs_2306_12867
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Wind Noise Reduction with a Diffusion-based Stochastic Regeneration Model
Lemercier, Jean-Marie
Thiemann, Joachim
Koning, Raphael
Gerkmann, Timo
Audio and Speech Processing
Machine Learning
Sound
In this paper we present a method for single-channel wind noise reduction using our previously proposed diffusion-based stochastic regeneration model combining predictive and generative modelling. We introduce a non-additive speech in noise model to account for the non-linear deformation of the membrane caused by the wind flow and possible clipping. We show that our stochastic regeneration model outperforms other neural-network-based wind noise reduction methods as well as purely predictive and generative models, on a dataset using simulated and real-recorded wind noise. We further show that the proposed method generalizes well by testing on an unseen dataset with real-recorded wind noise. Audio samples, data generation scripts and code for the proposed methods can be found online (https://uhh.de/inf-sp-storm-wind).
title Wind Noise Reduction with a Diffusion-based Stochastic Regeneration Model
topic Audio and Speech Processing
Machine Learning
Sound
url https://arxiv.org/abs/2306.12867