首页> 外文会议>2016 International Conference on Accessibility to Digital World >Transforming the vowel/diphthong formants of one Assamese variety to another: A GMM based approach
【24h】

Transforming the vowel/diphthong formants of one Assamese variety to another: A GMM based approach

机译:将一种阿萨姆语元音/双音共振峰转换为另一种:基于GMM的方法

获取原文
获取原文并翻译 | 示例

摘要

This study attempts to transform the formant frequencies of the vowels/diphthongs of one variety of the Assamese language to another variety, using a statistical approach based on Gaussian Mixture Models(GMMs). Formants are the resonant frequencies of the vocal tract and are responsible for the differences in quality among different periodic sounds. They can therefore be used to represent regional variations in the vowel and diphthong sounds. A GMM based mapping function is developed to map the most informative formants, F1, F2 & F3 of vowels/diphthongs of one variety of Assamese to the other. Objective evaluation results show that the mapping function is successful in bringing the vowel/diphthong formants of the source variety closer to the target variety.
机译:这项研究试图使用基于高斯混合模型(GMMs)的统计方法,将一种阿萨姆语的元音/双音的共振峰频率转换为另一种。共振峰是声道的共振频率,是造成不同周期性声音之间质量差异的原因。因此,它们可以用来表示元音和双峰音的区域变化。开发了基于GMM的映射功能,可以将多种阿萨姆语元音/双音的信息最丰富的共振峰F1,F2和F3映射到另一种。客观的评估结果表明,映射功能可以成功地使源变体的元音/双音共振峰更接近目标变体。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号