Prosodic Features Control by Symbols as Input of Sequence-to-Sequence Acoustic Modeling for Neural TTS
Summary :
This paper describes a method to control prosodic features using phonetic and prosodic symbols as input of attention-based sequence-to-sequence (seq2seq) acoustic modeling (AM) for neural text-to-speech (TTS). The method involves inserting a sequence of prosodic symbols between phonetic symbols that are then used to reproduce prosodic acoustic features, i.e. accents, pauses, accent breaks, and sentence endings, in several seq2seq AM methods. The proposed phonetic and prosodic labels have simple descriptions and a low production cost. By contrast, the labels of conventional statistical parametric speech synthesis methods are complicated, and the cost of time alignments such as aligning the boundaries of phonemes is high. The proposed method does not need the boundary positions of phonemes. We propose an automatic conversion method for conventional labels and show how to automatically reproduce pitch accents and phonemes. The results of objective and subjective evaluations show the effectiveness of our method.
- Publication
- IEICE TRANSACTIONS on Information Vol.E104-D No.2 pp.302-311
- Publication Date
- 2021/02/01
- Publicized
- 2020/11/09
- Online ISSN
- 1745-1361
- DOI
- 10.1587/transinf.2020EDP7104
- Type of Manuscript
- PAPER
- Category
- Speech and Hearing
Authors
Kiyoshi KURIHARA
NHK STRL
Nobumasa SEIYAMA
NHK STRL
Tadashi KUMANO
NHK STRL
Keyword
Latest Issue
Copyrights notice of machine-translated contents
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Cite this
Copy
Kiyoshi KURIHARA, Nobumasa SEIYAMA, Tadashi KUMANO, "Prosodic Features Control by Symbols as Input of Sequence-to-Sequence Acoustic Modeling for Neural TTS" in IEICE TRANSACTIONS on Information,
vol. E104-D, no. 2, pp. 302-311, February 2021, doi: 10.1587/transinf.2020EDP7104.
Abstract: This paper describes a method to control prosodic features using phonetic and prosodic symbols as input of attention-based sequence-to-sequence (seq2seq) acoustic modeling (AM) for neural text-to-speech (TTS). The method involves inserting a sequence of prosodic symbols between phonetic symbols that are then used to reproduce prosodic acoustic features, i.e. accents, pauses, accent breaks, and sentence endings, in several seq2seq AM methods. The proposed phonetic and prosodic labels have simple descriptions and a low production cost. By contrast, the labels of conventional statistical parametric speech synthesis methods are complicated, and the cost of time alignments such as aligning the boundaries of phonemes is high. The proposed method does not need the boundary positions of phonemes. We propose an automatic conversion method for conventional labels and show how to automatically reproduce pitch accents and phonemes. The results of objective and subjective evaluations show the effectiveness of our method.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.2020EDP7104/_p
Copy
@ARTICLE{e104-d_2_302,
author={Kiyoshi KURIHARA, Nobumasa SEIYAMA, Tadashi KUMANO, },
journal={IEICE TRANSACTIONS on Information},
title={Prosodic Features Control by Symbols as Input of Sequence-to-Sequence Acoustic Modeling for Neural TTS},
year={2021},
volume={E104-D},
number={2},
pages={302-311},
abstract={This paper describes a method to control prosodic features using phonetic and prosodic symbols as input of attention-based sequence-to-sequence (seq2seq) acoustic modeling (AM) for neural text-to-speech (TTS). The method involves inserting a sequence of prosodic symbols between phonetic symbols that are then used to reproduce prosodic acoustic features, i.e. accents, pauses, accent breaks, and sentence endings, in several seq2seq AM methods. The proposed phonetic and prosodic labels have simple descriptions and a low production cost. By contrast, the labels of conventional statistical parametric speech synthesis methods are complicated, and the cost of time alignments such as aligning the boundaries of phonemes is high. The proposed method does not need the boundary positions of phonemes. We propose an automatic conversion method for conventional labels and show how to automatically reproduce pitch accents and phonemes. The results of objective and subjective evaluations show the effectiveness of our method.},
keywords={},
doi={10.1587/transinf.2020EDP7104},
ISSN={1745-1361},
month={February},}
Copy
TY - JOUR
TI - Prosodic Features Control by Symbols as Input of Sequence-to-Sequence Acoustic Modeling for Neural TTS
T2 - IEICE TRANSACTIONS on Information
SP - 302
EP - 311
AU - Kiyoshi KURIHARA
AU - Nobumasa SEIYAMA
AU - Tadashi KUMANO
PY - 2021
DO - 10.1587/transinf.2020EDP7104
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E104-D
IS - 2
JA - IEICE TRANSACTIONS on Information
Y1 - February 2021
AB - This paper describes a method to control prosodic features using phonetic and prosodic symbols as input of attention-based sequence-to-sequence (seq2seq) acoustic modeling (AM) for neural text-to-speech (TTS). The method involves inserting a sequence of prosodic symbols between phonetic symbols that are then used to reproduce prosodic acoustic features, i.e. accents, pauses, accent breaks, and sentence endings, in several seq2seq AM methods. The proposed phonetic and prosodic labels have simple descriptions and a low production cost. By contrast, the labels of conventional statistical parametric speech synthesis methods are complicated, and the cost of time alignments such as aligning the boundaries of phonemes is high. The proposed method does not need the boundary positions of phonemes. We propose an automatic conversion method for conventional labels and show how to automatically reproduce pitch accents and phonemes. The results of objective and subjective evaluations show the effectiveness of our method.
ER -
English
