Requirements and Motivations of Low-Resource Speech Synthesis for Language Revitalization

Aidan Pine; Dan Wells; Nathan Thanyehténhas Brinklow; Patrick Littell; Korin Richmond

Requirements and Motivations of Low-Resource Speech Synthesis for Language Revitalization

Aidan Pine, Dan Wells, Nathan Thanyehténhas Brinklow, Patrick Littell, Korin Richmond

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract / Description of output

This paper describes the motivation and development of speech synthesis systems for the purposes of language revitalization. By building speech synthesis systems for three Indigenous languages spoken in Canada, Kanien’kéha, Gitksan & SENĆOŦEN, we re-evaluate the question of how much data is required to build low-resource speech synthesis systems featuring state-of-the-art neural models. For example, preliminary results with English data show that a FastSpeech2 model trained with 1 hour of training data can produce speech with comparable naturalness to a Tacotron2 model trained with 10 hours of data. Finally, we motivate future research in evaluation and classroom integration in the field of speech synthesis for language revitalization.

Original language	English
Title of host publication	Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics Volume 1: Long Papers
Editors	Smaranda Muresan, Preslav Nakov, Aline Villavicencio
Publisher	ACL Anthology
Pages	14
Number of pages	7346
Volume	1
ISBN (Electronic)	978-1-955917-21-6
Publication status	Published - 16 May 2022
Event	60th Annual Meeting of the Association for Computational Linguistics - The Convention Centre Dublin, Dublin, Ireland Duration: 22 May 2022 → 27 May 2022 https://www.2022.aclweb.org

Conference

Conference	60th Annual Meeting of the Association for Computational Linguistics
Abbreviated title	ACL 2022
Country/Territory	Ireland
City	Dublin
Period	22/05/22 → 27/05/22
Internet address	https://www.2022.aclweb.org

Access to Document

ACL_Speech_SynthesisAccepted author manuscript, 324 KB

https://aclanthology.org/2022.acl-long.507Licence: Creative Commons: Attribution (CC-BY)

Cite this

Pine, A., Wells, D., Brinklow, N. T., Littell, P., & Richmond, K. (2022). Requirements and Motivations of Low-Resource Speech Synthesis for Language Revitalization. In S. Muresan, P. Nakov, & A. Villavicencio (Eds.), Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics Volume 1: Long Papers (Vol. 1, pp. 14). ACL Anthology. https://aclanthology.org/2022.acl-long.507

@inproceedings{f6b92153f193499b9f2a432e1d8dc7ed,

title = "Requirements and Motivations of Low-Resource Speech Synthesis for Language Revitalization",

abstract = "This paper describes the motivation and development of speech synthesis systems for the purposes of language revitalization. By building speech synthesis systems for three Indigenous languages spoken in Canada, Kanien{\textquoteright}k{\'e}ha, Gitksan & SEN{\'C}OŦEN, we re-evaluate the question of how much data is required to build low-resource speech synthesis systems featuring state-of-the-art neural models. For example, preliminary results with English data show that a FastSpeech2 model trained with 1 hour of training data can produce speech with comparable naturalness to a Tacotron2 model trained with 10 hours of data. Finally, we motivate future research in evaluation and classroom integration in the field of speech synthesis for language revitalization.",

author = "Aidan Pine and Dan Wells and Brinklow, {Nathan Thanyeht{\'e}nhas} and Patrick Littell and Korin Richmond",

year = "2022",

month = may,

day = "16",

language = "English",

volume = "1",

pages = "14",

editor = "Smaranda Muresan and Preslav Nakov and Aline Villavicencio",

booktitle = "Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics Volume 1: Long Papers",

publisher = "ACL Anthology",

note = "60th Annual Meeting of the Association for Computational Linguistics, ACL 2022 ; Conference date: 22-05-2022 Through 27-05-2022",

url = "https://www.2022.aclweb.org",

}

Pine, A, Wells, D, Brinklow, NT, Littell, P & Richmond, K 2022, Requirements and Motivations of Low-Resource Speech Synthesis for Language Revitalization. in S Muresan, P Nakov & A Villavicencio (eds), Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics Volume 1: Long Papers. vol. 1, ACL Anthology, pp. 14, 60th Annual Meeting of the Association for Computational Linguistics, Dublin, Ireland, 22/05/22. <https://aclanthology.org/2022.acl-long.507>

Requirements and Motivations of Low-Resource Speech Synthesis for Language Revitalization. / Pine, Aidan; Wells, Dan; Brinklow, Nathan Thanyehténhas et al.
Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics Volume 1: Long Papers. ed. / Smaranda Muresan; Preslav Nakov; Aline Villavicencio. Vol. 1 ACL Anthology, 2022. p. 14.

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

TY - GEN

T1 - Requirements and Motivations of Low-Resource Speech Synthesis for Language Revitalization

AU - Pine, Aidan

AU - Wells, Dan

AU - Brinklow, Nathan Thanyehténhas

AU - Littell, Patrick

AU - Richmond, Korin

PY - 2022/5/16

Y1 - 2022/5/16

N2 - This paper describes the motivation and development of speech synthesis systems for the purposes of language revitalization. By building speech synthesis systems for three Indigenous languages spoken in Canada, Kanien’kéha, Gitksan & SENĆOŦEN, we re-evaluate the question of how much data is required to build low-resource speech synthesis systems featuring state-of-the-art neural models. For example, preliminary results with English data show that a FastSpeech2 model trained with 1 hour of training data can produce speech with comparable naturalness to a Tacotron2 model trained with 10 hours of data. Finally, we motivate future research in evaluation and classroom integration in the field of speech synthesis for language revitalization.

AB - This paper describes the motivation and development of speech synthesis systems for the purposes of language revitalization. By building speech synthesis systems for three Indigenous languages spoken in Canada, Kanien’kéha, Gitksan & SENĆOŦEN, we re-evaluate the question of how much data is required to build low-resource speech synthesis systems featuring state-of-the-art neural models. For example, preliminary results with English data show that a FastSpeech2 model trained with 1 hour of training data can produce speech with comparable naturalness to a Tacotron2 model trained with 10 hours of data. Finally, we motivate future research in evaluation and classroom integration in the field of speech synthesis for language revitalization.

M3 - Conference contribution

VL - 1

SP - 14

BT - Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics Volume 1: Long Papers

A2 - Muresan, Smaranda

A2 - Nakov, Preslav

A2 - Villavicencio, Aline

PB - ACL Anthology

T2 - 60th Annual Meeting of the Association for Computational Linguistics

Y2 - 22 May 2022 through 27 May 2022

ER -

Requirements and Motivations of Low-Resource Speech Synthesis for Language Revitalization

Abstract / Description of output

Conference

Access to Document

Fingerprint

Cite this