A preliminary study on cross-language knowledge transfer for low-resource Taiwanese Mandarin ASR

A preliminary study on cross-language knowledge transfer for low-resource Taiwanese Mandarin ASR
复制标题

DOI：
10.1109/icsda.2016.7918980
复制
发表时间：
2016-10
期刊：
2016 Conference of The Oriental Chapter of International Committee for Coordination and Standardization of Speech Databases and Assessment Techniques (O-COCOSDA)
影响因子：
--
通讯作者：
Chien-Ting Lin;Yih-Ru Wang;Sin-Horng Chen;Y. Liao
Chien-Ting Lin;Yih-Ru Wang;Sin-Horng Chen;Y. Liao
中科院分区：
其他
文献类型：
--
作者：
Chien-Ting Lin;Yih-Ru Wang;Sin-Horng Chen;Y. Liao

文献摘要

被引文献

相似文献

深度神经网络（DNN）是当今最先进的自动语音识别（ASR）技术。然而，成功的关键是需要大量目标语言的语音数据来很好地训练DNN。遗憾的是，在台湾只有少数的小型台语国语语料库。因此，在本文中，两种跨语言的知识转移方法进行评估，以建立一个高性能的台湾国语ASR包括（1）一个隐藏层和（2）共享隐藏层的方法。实验结果表明：（1）共享隐藏层的方法取得了最好的性能;（2）系统对不同的说话人和音素具有鲁棒性。
The deep neural networks (DNNs) are the state-of-the-art automatic speech recognition (ASR) technique nowadays. However, the key to success is that a large amount of speech data of the target language is required to well train DNNs. Unfortunately, there are only few small Taiwanese Mandarin Speech corpora available in Taiwan. Therefore, in this paper, two cross-language knowledge transfer approaches are evaluated for building a high performance Taiwanese Mandarin ASR including (1) a borrowed-hidden-layer and (2) a shared-hidden-layer method. Experimental results show that (1) the shared-hidden-layer method achieved the best performance and (2) the system is robustness to different speaker and phone.