匹配追踪说话人自适应方法 |
| |
引用本文: | 张文林,屈丹,李弼程.匹配追踪说话人自适应方法[J].声学学报,2014,39(4):523-530. |
| |
作者姓名: | 张文林 屈丹 李弼程 |
| |
作者单位: | 解放军信息工程大学信息系统工程学院 郑州 450002 |
| |
基金项目: | 国家自然科学基金(61175017);国家高技术研究发展计划(863)(2012AA011603)资助 |
| |
摘 要: | 针对现有子空间自适应方法无法确定最佳说话人子空间的问题,提出一种基于匹配追踪的说话人自适应方法。将说话人自适应视为一种高维信号的稀疏分解问题,利用本征音和参考说话人超矢量的各自优势联合构造说话人字典;依据匹配追踪原理,通过迭代优化,以后验方式确定最佳说话人子空间维数及其基矢量。引入冗余基矢量检测与去除机制以保证算法的稳定性,并通过快速递推算法得到新说话人坐标。基于汉语连续语音识别的有监督说话人自适应实验结果表明,与本征音及参考说话人加权方法相比,平均有调音节正识率相对提高了1.9%。
|
关 键 词: | 说话人 自适应方法 匹配追踪 连续语音识别 递推算法 稀疏分解 空间维数 子空间方法 语音识别系统 加权方法 |
收稿时间: | 2012-11-20 |
Speaker adaptation using matching pursuit |
| |
Institution: | Institute of Information System Engineering, PLA Information Engineering University Zhengzhou 450002 |
| |
Abstract: | Current speaker subspace based adaptation method cannot obtain the best speaker subspace. A speaker adaptation method based on matching pursuit was proposed to adress this problem. Speaker adaptation was viewed as the sparse decomposition of a high dimensional speaker supervector with an over-complete dictionary, which was constructed by eigenvoices and reference speaker supervectors. Through an efficient iteratively optimization process, the best speaker dependent subspace was determined in a maximum a posterior way. A redundant bases removing mechanism was introduced to ensure the numeric stability and new speaker's coordinate was obtained through a fast recurrence algorithm. Superised speaker adaptation on a Chinese continuous speech recognition system show that compared with the eigenvoice and reference speaker weighting methods, the recognition accuracy was improved by relatively 1.9% |
| |
Keywords: | |
本文献已被 CNKI 等数据库收录! |
| 点击此处可从《声学学报》浏览原始摘要信息 |
| 点击此处可从《声学学报》下载免费的PDF全文 |
|