中文标准女声音库(10000句)
Chinese Standard Mandarin Speech Copus(10000 Sentences)
本次开放的数据仅支持非商用! 问题反馈: sales@data-baker.com
SUPPORT NON-COMMERCIAL USE ONLY! CONTACT: sales@data-baker.com
语音合成是通过机械的、电子的方法产生人造语音的技术。TTS技术(又称文语转换技术)隶属于语音合成,它是将计算机自己产生的、或外部输入的文字信息转变为可以听得懂的、流利的口语输出的技术。
TTS语音合成技术是实现人机语音通信关键技术之一。使电脑具有类似于人一样的说话能力,是当今时代信息产业的重要竞争市场。和语音识别ASR相比,语音合成的技术相对来说要成熟一些,是应用范围较广的技术。
随着人工智能产业的飞速发展,语音合成系统也得到了更加广泛的应用。除了语音合成初期的清晰度、可懂度以外,人们对语音合成的自然度、节奏感以及音质的要求也越来越高。而语音库的质量也是决定语音合成效果的关键因素。
【中文标准女声音库】采集对象的音色风格知性阳光、亲切自然,专业标准普通话女声,听感乐观积极。录制环境为专业录音室和录音软件,录音环境和设备自始至终保持不变,录音环境的信噪比不低于35dB;单声道录音,用48KHz 16比特采样频率、PCM WAV格式。录音语料涵盖各类新闻、小说、科技、娱乐、对话等领域,语料设计综合语料样本量,力求在有限的语料数据量内,对音节音子、类型、音调、音连以及韵律等尽可能全面的覆盖。根据合成语音标注标准对音库进行文本音字校对、韵律层级标注、语音文件边界切分标注。
Speech synthesis is a technique that produces artificial speech by mechanical and electronic methods. Text-to-Speech service converts written text to natural-sounding speech to provide speech-synthesis capabilities for applications. Text-to-Speech service is a part of speech synthesis.
Text-to-Speech is one of the key technologies for realizing human-machine voice communication. It gives the computer the ability to speak like a human. Text-to-Speech functionality allows our characters to speak any text dynamically. Compared with Automatic Speech Recognition, Text-to-Speech is more mature and has a wider range of applications.
With the rapid development of the artificial intelligence industry, the TTS has also been used more widely. In addition to requiring the speech synthesis effect to be clear enough and understandable, people are increasingly demanding the naturalness, rhythm and sound quality of TTS speech synthesis. One of the key factors in determining the TTS synthesis effect is the quality of the speech corpus.