| 鉴别性最大后验概率线性回归说话人自适应研究 |
| |
| 引用本文： | 齐耀辉,潘复平,葛凤培,颜永红.鉴别性最大后验概率线性回归说话人自适应研究[J].北京理工大学学报,2015,35(9):946-950. |
| |
| 作者姓名： | 齐耀辉 潘复平 葛凤培 颜永红 |
| |
| 作者单位： | 北京理工大学信息与电子学院,北京100081;中国科学院声学研究所中国科学院语言声学与内容理解重点实验室,北京100190;河北师范大学物理科学与信息工程学院,河北,石家庄050024;中国科学院声学研究所中国科学院语言声学与内容理解重点实验室,北京,100190;北京理工大学信息与电子学院,北京100081;中国科学院声学研究所中国科学院语言声学与内容理解重点实验室,北京100190 |
| |
| 基金项目： | 国家自然科学基金资助项目(10925419,90920302,61072124,11074275,11161140319,91120001,61271426);中国科学院战略性先导科技专项(XDA06030100,XDA06030500);国家"八六三"计划项目(2012AA012503);中科院重点部署项目(KGZD-EW-103-2) |
| |
| 摘 要： | 为增强自适应后的声学模型的鉴别能力,提出了一种基于最大互信息(MMI)的鉴别性最大后验概率线性回归(MMI-DMAPLR)说话人自适应方法. 将最大互信息准则和最大后验概率(MAP)准则相结合,设计了一个新的目标函数来估计基于线性变换的自适应方法中的变换参数,在最大后验概率估计中加入了鉴别性. 大词汇量连续语音识别的实验结果表明,新方法在增强声学模型与测试数据的匹配性的同时,可以有效提高声学模型的鉴别能力,在少量自适应数据的情况下,其性能比最大后验概率线性回归(MAPLR)相对提高4.8%.
|
| 关 键 词： | 最大似然线性回归 最大后验概率线性回归 最大互信息 说话人自适应 |
| 收稿时间： | 2013/5/24 0:00:00 |
Investigation on Discriminative Maximum a Posteriori Linear Regression for Speaker Adaptation |
| |
| QI Yao-hui,PAN Fu-ping,GE Feng-pei and YAN Yong-hong.Investigation on Discriminative Maximum a Posteriori Linear Regression for Speaker Adaptation[J].Journal of Beijing Institute of Technology(Natural Science Edition),2015,35(9):946-950. |
| |
| Authors: | QI Yao-hui PAN Fu-ping GE Feng-pei and YAN Yong-hong |
| |
| Institution: | School of Information and Electronics, Beijing Institute of Technology, Beijing 100081, China;Key Laboratory of Speech Acoustics and Content Understanding, Institute of Acoustic, Chinese Academy of Science, Beijing 100190, China;College of Physics and Information Engineering, Hebei Normal University, Shijiazhuang, Hebei 050024, China,Key Laboratory of Speech Acoustics and Content Understanding, Institute of Acoustic, Chinese Academy of Science, Beijing 100190, China,Key Laboratory of Speech Acoustics and Content Understanding, Institute of Acoustic, Chinese Academy of Science, Beijing 100190, China and School of Information and Electronics, Beijing Institute of Technology, Beijing 100081, China;Key Laboratory of Speech Acoustics and Content Understanding, Institute of Acoustic, Chinese Academy of Science, Beijing 100190, China |
| |
| Abstract: | In order to increase the discriminative capability of the adapted acoustic model, the maximum mutual information based discriminative maximum a posteriori linear regression (MMI-DMAPLR) adaptation method was proposed. Combining the maximum mutual information criterion with maximum a posteriori (MAP) criterion, a new objective function was designed to estimate the transform parameters of adaptation method based on the linear transformation, to increase the discriminative capability in maximum a posteriori estimation. The experimental results in large vocabulary continuous recognition show that the proposed method can both enhance the match degree between the acoustic model and the test data and the discriminative power of acoustic model. Compared with maximum a posteriori linear regression (MAPLR), the proposed method can obtain 4.8% relative reduction in word error rate when the amount of data is limited. |
| |
| Keywords: | maximum likelihood linear regression maximum a posteriori linear regression maximum mutual information speaker adaptation |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
| 点击此处可从《北京理工大学学报》浏览原始摘要信息 |
| 点击此处可从《北京理工大学学报》下载免费的PDF全文 |
|
