Skip to content

Latest commit

 

History

History
128 lines (105 loc) · 5.12 KB

data.md

File metadata and controls

128 lines (105 loc) · 5.12 KB
Raw

Forced-aligned Speech Datasets

Here we release TextGrids for several datasets that have been forced aligned with Charsiu Forced Aligner. Hoepfully they might be helpful for your research. Forced alignment does not generate perfect alignments. Use at you own discrection.

English
Mandarin

Please cite this if you use these alignments in your research projects.

@article{zhu2022charsiu,
  title={Phone-to-audio alignment without text: A Semi-supervised Approach},
  author={Zhu, Jian and Zhang, Cong and Jurgens, David},
  journal={IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)},
  year={2022},
  url=https://arxiv.org/abs/2110.03876,
 }

Alignments for English datasets

Textgrids

You can find all textgrids for the trianing set (~860k utterances) in this google drive folder. It contains phone- and word-level alignments for the English subset of Common Voice (~2000 hours). It was aligned with charsiu_forced_aligner with the model charsiu/en_w2v2_fc_10ms. All filenames are matched. Only a few mismatched samples were discarded.

Audio

You can find the dataset at the Common Voice Project.

The audio data can also be easily accessed through the Common Voce page at HuggingFace hub. An account is needed for authentication.

from datasets import load_dataset

dataset = load_dataset("mozilla-foundation/common_voice_8_0", "en", split='train',use_auth_token=True)

Note that ~80GB of memory is needed to load the dataset into memory.

Please cite Common Voice if you use this dataset.

@inproceedings{commonvoice:2020,
 author = {Ardila, R. and Branson, M. and Davis, K. and Henretty, M. and Kohler, M. and Meyer, J. and Morais, R. and Saunders, L. and Tyers, F. M. and Weber, G.},
 title = {Common Voice: A Massively-Multilingual Speech Corpus},
 booktitle = {Proceedings of the 12th Conference on Language Resources and Evaluation (LREC 2020)},
 pages = {4211--4215},
 year = 2020
}

The grapheme-to-phoneme conversion was done automatically with g2p_en.

@misc{g2pE2019,
  author = {Park, Kyubyong & Kim, Jongseok},
  title = {g2pE},
  year = {2019},
  publisher = {GitHub},
  journal = {GitHub repository},
  howpublished = {\url{https://github.com/Kyubyong/g2p}}
}

Alignments for Mandarin Speech datasets

This repository contains phone- and word-level alignments for multiple Mandarin Chinese speech datasets, including MagicData (~755 hours), Aishell-1 (~150 hours), STCMDS (~100 hours), Datatang (~200 hours), THCHS-30 (~30 hours) and PrimeWords (~100 hours).

Textgrids

You can download all textgrids here. The forced alignment was done with charsiu_forced_aligner using model charsiu/zh_xlsr_fc_10ms. Only Praat textgrid files are distributed. Sentences with Englist letters and numbers were all removed. Misaligned files were also discarded.

The grapheme-to-phoneme conversion was done automatically with g2pM.

@article{park2020g2pm,
 author={Park, Kyubyong and Lee, Seanie},
 title = {A Neural Grapheme-to-Phoneme Conversion Package for Mandarin Chinese Based on a New Open Benchmark Dataset
},
 journal={Proc. Interspeech 2020},
 url = {https://arxiv.org/abs/2004.03136},
 year = {2020}
}

Audio data

The original audio data can be downloaded via OpenSLR. All filenames are matched. Please also cite the original datasets.

Aishell-1 (~150 hours): https://openslr.org/33/

@inproceedings{aishell_2017,
  title={AIShell-1: An Open-Source Mandarin Speech Corpus and A Speech Recognition Baseline},
  author={Hui Bu, Jiayu Du, Xingyu Na, Bengu Wu, Hao Zheng},
  booktitle={Oriental COCOSDA 2017},
  pages={Submitted},
  year={2017}
}

MagicData (~755 hours): https://openslr.org/68/

Please cite the corpus as "Magic Data Technology Co., Ltd., "http://www.imagicdatatech.com/index.php/home/dataopensource/data_info/id/101", 05/2019".

Datatang (~200 hours): https://openslr.org/62/

Please cite the corpus as “aidatatang_200zh, a free Chinese Mandarin speech corpus by Beijing DataTang Technology Co., Ltd ( www.datatang.com )”.

STCMDS (~100 hours): https://openslr.org/38/

Please cite the data as “ST-CMDS-20170001_1, Free ST Chinese Mandarin Corpus”.

PrimeWords (~100 hours): https://openslr.org/47/

  @misc{primewords_201801,
    title={Primewords Chinese Corpus Set 1},
    author={Primewords Information Technology Co., Ltd.},
    year={2018},
    note={\url{https://www.primewords.cn}}
    }

THCHS-30 (~30 hours): https://openslr.org/18/

@misc{THCHS30_2015,
  title={THCHS-30 : A Free Chinese Speech Corpus},
  author={Dong Wang, Xuewei Zhang, Zhiyong Zhang},
  year={2015},
  url={http://arxiv.org/abs/1512.01882}
}