Hi-Fi Multi-Speaker English TTS Dataset

被引：30

作者：

Bakhturina, Evelina ^{[1
]}

Lavrukhin, Vitaly ^{[1
]}

Ginsburg, Boris ^{[1
]}

Zhang, Yang ^{[1
]}

机构：

[1] NVIDIA, Santa Clara, CA 95051 USA

来源：

INTERSPEECH 2021 | 2021年

关键词：

D O I：

10.21437/Interspeech.2021-1599

中图分类号：

R36 [病理学]; R76 [耳鼻咽喉科学];

学科分类号：

100104 ; 100213 ;

摘要：

This paper introduces a new multi-speaker English dataset for training text-to-speech models. The dataset is based on LibriVox audiobooks and Project Gutenberg texts, both in the public domain. The new dataset contains about 292 hours of speech from 10 speakers with at least 17 hours per speaker sampled at 44.1 kHz. To select speech samples with high quality, we considered audio recordings with a signal bandwidth of at least 13 kHz and a signal-to-noise ratio (SNR) of at least 32 dB. The dataset is publicly released at "http://www.openslr.org/109/".

引用

页码：2776 / 2780

页数：5

共 34 条

[1]

[Anonymous], Librivox: Free Public Domain Audiobooks

[2]

Ardila R., 2019, ARXIV191206670

[3]

Beliaev S., 2020, ARXIV200505514

[4]

Chen M., 2020, ARXIV200604664

[5]

Chen Mingjian, 2021, ARXIV210300993

[6]

Cieri C., 2004, LREC

[7]

Godfrey J., 1992, ICASSP

[8]

Graves Alex, 2006, P 23 INT C MACH LEAR, DOI [DOI 10.1145/1143844.1143891, 10.1145/1143844.1143891]

[9] Violin plots: A box plot-density trace synergism [J].

Hintze, JL ;

Nelson, RD .

AMERICAN STATISTICIAN, 1998, 52 (02) :181-184

[10]

Ito Keith, 2017, The lj speech dataset

← 1 2 3 4 →