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(OpenSource Name):marytts/gradle-marytts-voicebuilding-plugin开源软件地址(OpenSource Url):https://github.com/marytts/gradle-marytts-voicebuilding-plugin开源编程语言(OpenSource Language):Groovy 82.6%开源软件介绍(OpenSource Introduction):Gradle MaryTTS voicebuilding pluginUse this plugin to build new unitselection voices for MaryTTS. It's a replacement for the legacy VoiceImportTools. Applying the pluginAdd this at the top of your plugins {
id 'de.dfki.mary.voicebuilding-legacy' version '5.4'
} For details, see https://plugins.gradle.org/plugin/de.dfki.mary.voicebuilding-legacy. Note that Gradle v6.2 or higher is required. PrerequisitesRequired third-party softwareJava 8 or higher is required. SoX and Edinburgh Speech Tools must be installed and on the Mac OSXWith Homebrew, just run
LinuxOn Debian-based systems, just run
Project layoutIn your project directory, place the source audio, text, and label files under your
The audio location can be easily customized, by configuring the layout in the // wav files under "$projectDir/wav"
wav.srcDir = file('wav') Configuring the voiceCrucial details of the voice to be built are configured in the marytts {
voice {
name = 'my_voice'
gender = 'female'
language = 'en'
region = 'US'
domain = 'general'
type = 'unit selection'
description = 'A female English unit selection voice'
samplingRate = 16000
license {
name = 'Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International'
shortName = 'CC BY-NC-SA 4.0'
url = 'http://creativecommons.org/licenses/by-nc-sa/4.0/'
}
}
} Any of these can (and should) be changed as needed, according to the dataset used to build the voice.
The default values do not need to be specified, and the Building the voiceTo assemble and test the voice, run
Afterwards, the packaged voice component (and its XML descriptor) will be found under |
2023-10-27
2022-08-15
2022-08-17
2022-09-23
2022-08-13
请发表评论