Remix-cycle-consistent Learning on Adversarially Learned Separator for Accurate and Stable Unsupervised Speech Separation

03/26/2022
by   Kohei Saijo, et al.
0

A new learning algorithm for speech separation networks is designed to explicitly reduce residual noise and artifacts in the separated signal in an unsupervised manner. Generative adversarial networks are known to be effective in constructing separation networks when the ground truth for the observed signal is inaccessible. Still, weak objectives aimed at distribution-to-distribution mapping make the learning unstable and limit their performance. This study introduces the remix-cycle-consistency loss as a more appropriate objective function and uses it to fine-tune adversarially learned source separation models. The remix-cycle-consistency loss is defined as the difference between the mixed speech observed at microphones and the pseudo-mixed speech obtained by alternating the process of separating the mixed sound and remixing its outputs with another combination. The minimization of this loss leads to an explicit reduction in the distortions in the output of the separation network. Experimental comparisons with multichannel speech separation demonstrated that the proposed method achieved high separation accuracy and learning stability comparable to supervised learning.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
04/23/2022

Heterogeneous Separation Consistency Training for Adaptation of Unsupervised Speech Separation

Recently, supervised speech separation has made great progress. However,...
research
11/11/2019

Unsupervised Training for Deep Speech Source Separation with Kullback-Leibler Divergence Based Probabilistic Loss Function

In this paper, we propose a multi-channel speech source separation with ...
research
10/22/2020

DBNET: DOA-driven beamforming network for end-to-end farfield sound source separation

Many deep learning techniques are available to perform source separation...
research
03/10/2023

Distribution Preserving Source Separation With Time Frequency Predictive Models

We provide an example of a distribution preserving source separation met...
research
01/25/2023

Separate And Diffuse: Using a Pretrained Diffusion Model for Improving Source Separation

The problem of speech separation, also known as the cocktail party probl...
research
10/12/2022

Individualized Conditioning and Negative Distances for Speaker Separation

Speaker separation aims to extract multiple voices from a mixed signal. ...
research
12/10/2018

A Computationally Efficient and Practically Feasible Two Microphones Blind Speech Separation Method

Traditionally, Blind Speech Separation techniques are computationally ex...

Please sign up or login with your details

Forgot password? Click here to reset