U-Vectors: Generating Clusterable Speaker Embedding from Unlabeled Data

Speaker recognition deals with recognizing speakers by their speech. Most speaker recognition systems are built upon two stages, the first stage extracts low dimensional correlation embeddings from speech, and the second performs the classification task. The robustness of a speaker recognition syste...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Muhammad Firoz Mridha, Abu Quwsar Ohi, Muhammad Mostafa Monowar, Md. Abdul Hamid, Md. Rashedul Islam, Yutaka Watanobe
Formato:	article
Lenguaje:	EN
Publicado:	MDPI AG 2021
Materias:	speaker recognition clustering twin networks deep learning Technology T Engineering (General). Civil engineering (General) TA1-2040 Biology (General) QH301-705.5 Physics QC1-999 Chemistry QD1-999
Acceso en línea:	https://doaj.org/article/8794b082046b46ef882b1427831a56b0
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:8794b082046b46ef882b1427831a56b0
record_format	dspace
spelling	oai:doaj.org-article:8794b082046b46ef882b1427831a56b02021-11-11T15:08:50ZU-Vectors: Generating Clusterable Speaker Embedding from Unlabeled Data10.3390/app1121100792076-3417https://doaj.org/article/8794b082046b46ef882b1427831a56b02021-10-01T00:00:00Zhttps://www.mdpi.com/2076-3417/11/21/10079https://doaj.org/toc/2076-3417Speaker recognition deals with recognizing speakers by their speech. Most speaker recognition systems are built upon two stages, the first stage extracts low dimensional correlation embeddings from speech, and the second performs the classification task. The robustness of a speaker recognition system mainly depends on the extraction process of speech embeddings, which are primarily pre-trained on a large-scale dataset. As the embedding systems are pre-trained, the performance of speaker recognition models greatly depends on domain adaptation policy, which may reduce if trained using inadequate data. This paper introduces a speaker recognition strategy dealing with unlabeled data, which generates clusterable embedding vectors from small fixed-size speech frames. The unsupervised training strategy involves an assumption that a small speech segment should include a single speaker. Depending on such a belief, a pairwise constraint is constructed with noise augmentation policies, used to train AutoEmbedder architecture that generates speaker embeddings. Without relying on domain adaption policy, the process unsupervisely produces clusterable speaker embeddings, termed unsupervised vectors (u-vectors). The evaluation is concluded in two popular speaker recognition datasets for English language, TIMIT, and LibriSpeech. Also, a Bengali dataset is included to illustrate the diversity of the domain shifts for speaker recognition systems. Finally, we conclude that the proposed approach achieves satisfactory performance using pairwise architectures.Muhammad Firoz MridhaAbu Quwsar OhiMuhammad Mostafa MonowarMd. Abdul HamidMd. Rashedul IslamYutaka WatanobeMDPI AGarticlespeaker recognitionclusteringtwin networksdeep learningTechnologyTEngineering (General). Civil engineering (General)TA1-2040Biology (General)QH301-705.5PhysicsQC1-999ChemistryQD1-999ENApplied Sciences, Vol 11, Iss 10079, p 10079 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	speaker recognition clustering twin networks deep learning Technology T Engineering (General). Civil engineering (General) TA1-2040 Biology (General) QH301-705.5 Physics QC1-999 Chemistry QD1-999
spellingShingle	speaker recognition clustering twin networks deep learning Technology T Engineering (General). Civil engineering (General) TA1-2040 Biology (General) QH301-705.5 Physics QC1-999 Chemistry QD1-999 Muhammad Firoz Mridha Abu Quwsar Ohi Muhammad Mostafa Monowar Md. Abdul Hamid Md. Rashedul Islam Yutaka Watanobe U-Vectors: Generating Clusterable Speaker Embedding from Unlabeled Data
description	Speaker recognition deals with recognizing speakers by their speech. Most speaker recognition systems are built upon two stages, the first stage extracts low dimensional correlation embeddings from speech, and the second performs the classification task. The robustness of a speaker recognition system mainly depends on the extraction process of speech embeddings, which are primarily pre-trained on a large-scale dataset. As the embedding systems are pre-trained, the performance of speaker recognition models greatly depends on domain adaptation policy, which may reduce if trained using inadequate data. This paper introduces a speaker recognition strategy dealing with unlabeled data, which generates clusterable embedding vectors from small fixed-size speech frames. The unsupervised training strategy involves an assumption that a small speech segment should include a single speaker. Depending on such a belief, a pairwise constraint is constructed with noise augmentation policies, used to train AutoEmbedder architecture that generates speaker embeddings. Without relying on domain adaption policy, the process unsupervisely produces clusterable speaker embeddings, termed unsupervised vectors (u-vectors). The evaluation is concluded in two popular speaker recognition datasets for English language, TIMIT, and LibriSpeech. Also, a Bengali dataset is included to illustrate the diversity of the domain shifts for speaker recognition systems. Finally, we conclude that the proposed approach achieves satisfactory performance using pairwise architectures.
format	article
author	Muhammad Firoz Mridha Abu Quwsar Ohi Muhammad Mostafa Monowar Md. Abdul Hamid Md. Rashedul Islam Yutaka Watanobe
author_facet	Muhammad Firoz Mridha Abu Quwsar Ohi Muhammad Mostafa Monowar Md. Abdul Hamid Md. Rashedul Islam Yutaka Watanobe
author_sort	Muhammad Firoz Mridha
title	U-Vectors: Generating Clusterable Speaker Embedding from Unlabeled Data
title_short	U-Vectors: Generating Clusterable Speaker Embedding from Unlabeled Data
title_full	U-Vectors: Generating Clusterable Speaker Embedding from Unlabeled Data
title_fullStr	U-Vectors: Generating Clusterable Speaker Embedding from Unlabeled Data
title_full_unstemmed	U-Vectors: Generating Clusterable Speaker Embedding from Unlabeled Data
title_sort	u-vectors: generating clusterable speaker embedding from unlabeled data
publisher	MDPI AG
publishDate	2021
url	https://doaj.org/article/8794b082046b46ef882b1427831a56b0
work_keys_str_mv	AT muhammadfirozmridha uvectorsgeneratingclusterablespeakerembeddingfromunlabeleddata AT abuquwsarohi uvectorsgeneratingclusterablespeakerembeddingfromunlabeleddata AT muhammadmostafamonowar uvectorsgeneratingclusterablespeakerembeddingfromunlabeleddata AT mdabdulhamid uvectorsgeneratingclusterablespeakerembeddingfromunlabeleddata AT mdrashedulislam uvectorsgeneratingclusterablespeakerembeddingfromunlabeleddata AT yutakawatanobe uvectorsgeneratingclusterablespeakerembeddingfromunlabeleddata
_version_	1718437154572468224

U-Vectors: Generating Clusterable Speaker Embedding from Unlabeled Data

Ejemplares similares