State-of-the-Art Analysis of Deep Learning-Based Monaural Speech Source Separation Techniques
The monaural speech source separation problem is an important application in the signal processing field. But recent interaction of deep learning algorithms with signal processing achieves remarkable performance improvement for speech source separation problems. This paper explores the numerous stat...
Main Authors: | , , |
---|---|
Format: | Article |
Language: | English |
Published: |
IEEE
2023-01-01
|
Series: | IEEE Access |
Subjects: | |
Online Access: | https://ieeexplore.ieee.org/document/10011434/ |
_version_ | 1797902199837163520 |
---|---|
author | Swati Soni Ram Narayan Yadav Lalita Gupta |
author_facet | Swati Soni Ram Narayan Yadav Lalita Gupta |
author_sort | Swati Soni |
collection | DOAJ |
description | The monaural speech source separation problem is an important application in the signal processing field. But recent interaction of deep learning algorithms with signal processing achieves remarkable performance improvement for speech source separation problems. This paper explores the numerous state-of-the-art deep learning-based monaural speech source separation algorithms in the time-frequency (T-F), time, and hybrid domains. The motivation, algorithm, and framework of different deep learning models for monaural speech source separation are analyzed. The benchmarked algorithms in the T-F domain can be categorized as deep neural networks (DNN), clustering, permutation, multi-task learning, computational auditory sense analysis (CASA), and phase reconstruction-based techniques, whereas the state-of-the-art time-domain approaches can be categorized as CNN, RNN, multi-scale fusion (MSF), and transformer-based techniques. The end-to-end post filter (E2EPF) is a hybrid algorithm combining T-F and time-domain works to achieve enhanced results. Time-domain models have shown improvement in separation performance compared to the T-F and hybrid domain models with small model sizes. Methods in T-F, time, and hybrid domains are compared using <inline-formula> <tex-math notation="LaTeX">$SDR$ </tex-math></inline-formula>, <inline-formula> <tex-math notation="LaTeX">$SI-SDR$ </tex-math></inline-formula>, <inline-formula> <tex-math notation="LaTeX">$SI-SNR$ </tex-math></inline-formula>, PESQ, and <inline-formula> <tex-math notation="LaTeX">$STOI$ </tex-math></inline-formula> as quality assessment metrics on some benchmark datasets. |
first_indexed | 2024-04-10T09:13:56Z |
format | Article |
id | doaj.art-1f864087e17e4616aa352e49091ecca3 |
institution | Directory Open Access Journal |
issn | 2169-3536 |
language | English |
last_indexed | 2024-04-10T09:13:56Z |
publishDate | 2023-01-01 |
publisher | IEEE |
record_format | Article |
series | IEEE Access |
spelling | doaj.art-1f864087e17e4616aa352e49091ecca32023-02-21T00:02:30ZengIEEEIEEE Access2169-35362023-01-01114242426910.1109/ACCESS.2023.323501010011434State-of-the-Art Analysis of Deep Learning-Based Monaural Speech Source Separation TechniquesSwati Soni0https://orcid.org/0000-0003-1147-3877Ram Narayan Yadav1Lalita Gupta2Department of Electronics and Communication Engineering, Maulana Azad National Institute of Technology, Bhopal, Madhya Pradesh, IndiaDepartment of Electronics and Communication Engineering, Maulana Azad National Institute of Technology, Bhopal, Madhya Pradesh, IndiaDepartment of Electronics and Communication Engineering, Maulana Azad National Institute of Technology, Bhopal, Madhya Pradesh, IndiaThe monaural speech source separation problem is an important application in the signal processing field. But recent interaction of deep learning algorithms with signal processing achieves remarkable performance improvement for speech source separation problems. This paper explores the numerous state-of-the-art deep learning-based monaural speech source separation algorithms in the time-frequency (T-F), time, and hybrid domains. The motivation, algorithm, and framework of different deep learning models for monaural speech source separation are analyzed. The benchmarked algorithms in the T-F domain can be categorized as deep neural networks (DNN), clustering, permutation, multi-task learning, computational auditory sense analysis (CASA), and phase reconstruction-based techniques, whereas the state-of-the-art time-domain approaches can be categorized as CNN, RNN, multi-scale fusion (MSF), and transformer-based techniques. The end-to-end post filter (E2EPF) is a hybrid algorithm combining T-F and time-domain works to achieve enhanced results. Time-domain models have shown improvement in separation performance compared to the T-F and hybrid domain models with small model sizes. Methods in T-F, time, and hybrid domains are compared using <inline-formula> <tex-math notation="LaTeX">$SDR$ </tex-math></inline-formula>, <inline-formula> <tex-math notation="LaTeX">$SI-SDR$ </tex-math></inline-formula>, <inline-formula> <tex-math notation="LaTeX">$SI-SNR$ </tex-math></inline-formula>, PESQ, and <inline-formula> <tex-math notation="LaTeX">$STOI$ </tex-math></inline-formula> as quality assessment metrics on some benchmark datasets.https://ieeexplore.ieee.org/document/10011434/Deep-clusteringdeep learningmonaural speech source separationpermutation invariant trainingtime domain speaker separation |
spellingShingle | Swati Soni Ram Narayan Yadav Lalita Gupta State-of-the-Art Analysis of Deep Learning-Based Monaural Speech Source Separation Techniques IEEE Access Deep-clustering deep learning monaural speech source separation permutation invariant training time domain speaker separation |
title | State-of-the-Art Analysis of Deep Learning-Based Monaural Speech Source Separation Techniques |
title_full | State-of-the-Art Analysis of Deep Learning-Based Monaural Speech Source Separation Techniques |
title_fullStr | State-of-the-Art Analysis of Deep Learning-Based Monaural Speech Source Separation Techniques |
title_full_unstemmed | State-of-the-Art Analysis of Deep Learning-Based Monaural Speech Source Separation Techniques |
title_short | State-of-the-Art Analysis of Deep Learning-Based Monaural Speech Source Separation Techniques |
title_sort | state of the art analysis of deep learning based monaural speech source separation techniques |
topic | Deep-clustering deep learning monaural speech source separation permutation invariant training time domain speaker separation |
url | https://ieeexplore.ieee.org/document/10011434/ |
work_keys_str_mv | AT swatisoni stateoftheartanalysisofdeeplearningbasedmonauralspeechsourceseparationtechniques AT ramnarayanyadav stateoftheartanalysisofdeeplearningbasedmonauralspeechsourceseparationtechniques AT lalitagupta stateoftheartanalysisofdeeplearningbasedmonauralspeechsourceseparationtechniques |