关于瓶颈特征藏语拉萨话连续语音识别研究1.PDFVIP

基于瓶颈特征的藏语拉萨话连续语音识别研究1 周楠 1 赵悦 1, † 李要嫱 1 徐晓娜 1 才旺拉姆 1 吴立成 1 1. 中央民族大学 信息工程学院，北京 100081； † 通讯作者，E-mail: zhaoyueso@muc.edu.cn 摘要 目前深度神经网络在语音识别中得到了广泛的应用，虽然其具有较高的稳健性和语义区分性，但其后验 特征无法用于成熟的高斯混合-隐马尔科夫声学建模框架中。然而，具有狭窄瓶颈的神经网络可以解决此问题， 其提取的瓶颈特征不仅具有语音长时相关性和紧凑表示的特点，并且可以代替传统的 MFCC 特征进行 GMM-HMM 声学建模。本文研究将瓶颈特征及其与MFCC 的复合特征用于藏语连续语音识别任务中。实验表 明，瓶颈复合特征取得了较深度神经网络后验特征和单瓶颈特征更好的识别表现。 关键词 藏语拉萨话；连续语音识别；高斯混合-隐马尔科夫模型；瓶颈特征；深度神经网络 中图分类号 TP391 Study on Continuous Speech Recognition based on Bottleneck Features for Lhasa-Tibetan Dialect 1 1, † 1 1 1 1 ZHOU Nan ，ZHAO Yue ，LI Yaoqiang ，XU Xiaona ，CAIWANG Lamu ，WU Licheng 1. School of Information Engineering ，Minzu University of China ， Beijing ，100081，China ; †Corresponding Author, E-mail: zhaoyueso@muc.edu.cn Abstract At present, deep neural network has been widely used in speech recognition, although it has high robustness and semantic distinction, but its posterior features cannot be used for GMM-HMM acoustic modeling framework. However, the neural network with a narrow bottleneck can solve this problem, and its bottleneck features not only have long term context-dependence and compact representation of speech signal, but also can replace the traditional MFCC features for GMM-HMM acoustic modeling. In this paper, we study on applying bottleneck features and its concatenated features with MFCC into Lhasa-Tibetan continuous speech recognition. The experimental results show that the concatenated features of bottleneck features and MFCC ac