Developing a Speech Corpus for Sinhala Speech Recognition

Developing a Speech Corpus for Sinhala Speech Recognition
复制标题

开发僧伽罗语语音识别语音语料库

DOI：
--
复制
发表时间：
2013
期刊：
影响因子：
--
通讯作者：
R. Weerasinghe
R. Weerasinghe
中科院分区：
--
文献类型：
--
作者：
Thilini Nadungodage;V. Welgama;R. Weerasinghe

文献摘要

被引文献

相似文献

语音语料库是基于统计模型的语音识别研究的主要组成部分，并且很大程度上影响了语音识别器的表现。应该为低资源的语言（例如僧伽罗语）建立良好的语音语料库。僧伽罗语言缺乏适当的语音语料库来进行语音识别研究。在本文中，我们介绍了为僧伽罗语设计设计和开发语音语料库的努力，因为这将促进对Sin-Hala语音识别的未来研究。
Speech corpora is a main part of statistical model based speech recognition research and highly affect the performance of a speech recognizer. Some effort should be given to build a good speech corpus for low-resourced languages such as Sinhala. Sinhala language suffers from the lack of proper speech corpora for speech recognition research. In this paper we present our effort on designing and developing a speech corpus for the Sinhala language as it would facilitate future research on Sin-hala speech recognition.