在线时间:8:00-16:00
迪恩网络APP
随时随地掌握行业动态
扫描二维码
关注迪恩网络微信公众号
开源软件名称:Jackiexiao/MTTS开源软件地址:https://github.com/Jackiexiao/MTTS开源编程语言:Python 97.4%开源软件介绍:本项目已停止维护,已相当老旧推荐:
欢迎加入
A Demo of MTTS Mandarin/Chinese Text to Speech FrontEndMandarin/Chinese Text to Speech based on statistical parametric speech synthesis using merlin toolkit 这只是一个语音合成前端的Demo,没有提供文本正则化,韵律预测功能,文字转拼音使用pypinyin,分词使用结巴分词,这两者的准确度也达不到商用水平。 其他语音合成项目传送门,端到端是不错的方向,自然度要优于merlin。 This is only a demo of mandarin frontend which is lack of some parts like "text normalization" and "prosody prediction", and the phone set && Question Set this project use havn't fully tested yet. 一个粗略的文档:A draft documentation written in Mandarin DataThere is no open-source mandarin speech synthesis dataset on the internet, this proj used thchs30 dataset to demostrate speech synthesis UPDATE open-source mandarin speech synthesis data from data-banker company, 开源的中文语音合成数据,感谢标贝公司 【数据下载】https://weixinxcxdb.oss-cn-beijing.aliyuncs.com/gwYinPinKu/BZNSYP.rar 【数据说明】http://www.data-baker.com/open_source.html Generated SamplesListen to https://jackiexiao.github.io/MTTS/ How To Reproduce
Context related annotation & Question SetInstallPython : python3.6
Run
Run Demo
Usage1. Generate HTS Label by wav and text
txtfile example
wav_directory example(Sampleing Rate should larger than 16khz)
2. Generate HTS Label by text with or without alignment file
see source
code for more information, but pay attention to the alignment file(sfs file), the format is 3. Forced-alignmentThis project use Montreal-Forced-Aligner to do forced alignment, if you want to get a better alignment, use your data to train a alignment-model, see mfa: algin-using-only-the-dataset
Prosody MarkYou can generate HTS Label without prosody mark. we assume that word segment is smaller than prosodic word(which is adjusted in code) "#0","#1", "#2","#3" and "#4" are the prosody labeling symbols.
Improvement to be done in future
Contributor
|
2023-10-27
2022-08-15
2022-08-17
2022-09-23
2022-08-13
请发表评论