%0 Journal Article
%T 应用于短时语音语种识别的时长扩展方法<br>Expanding the length of short utterances for short-duration language recognition
%A 苗晓晓
%A 张健
%A 索宏彬
%A 周若华
%A 颜永红
%J 清华大学学报（自然科学版）
%D 2018
%R 10.16511/j.cnki.qhdxxb.2018.25.015
%X 为解决待识别语音时长小于10 s时，语种识别性能急剧下降的问题，该文提出应用语音时域伸缩（time-scale modification，TSM）技术改变语音的长度（从而改变了语速），并保持其他频域信息不变。首先，对一段待识别语音，应用TSM技术转换为多条时域压缩和时域拉伸后的语音；其次，将这些不同语速的语音与原语音拼接起来，生成一个时长较长的语音；最后，送入语种识别系统进行识别。实验结果表明：所提出的语音时长扩展算法可以显著提升短时语音的语种识别性能。<br>Abstract：The language recognition (LR) accuracy is often significantly reduced when the test utterance duration is as short as 10 s or less. This paper describes a method to extend the utterance length using time-scale modification (TSM) which changes the speech rate without changing the spectral information. The algorithm first converts an utterance to several time-stretched or time-compressed versions using TSM. These modified versions with different speech rates are concatenated together with the original one to form a long-duration signal, which is subsequently fed into the LR system. Tests demonstrate that this duration modification method dramatically improves the performance for short utterances.
%K 语种识别
%K 短时
%K 时域伸缩
%K 语速
%K &lt
%K br&gt
%K language recognition
%K short-duration
%K time-scale modification
%K speech rate
%U http://jst.tsinghuajournals.com/CN/Y2018/V58/I3/254