Learning Speech Rate in Speech Recognition

被引:0
作者
Zeng, Xiangyu [1 ,3 ]
Yin, Shi [1 ,4 ]
Wang, Dong [1 ,2 ]
机构
[1] Tsinghua Univ, Ctr Speech & Language Technol CSLT, Res Inst Informat Technol, Beijing, Peoples R China
[2] Tsinghua Natl Lab Informat Sci & Technol, Beijing, Peoples R China
[3] Beijing Univ Posts & Telecommun, Beijing, Peoples R China
[4] Chongqing Univ Posts & Telecommun, Chongqing, Peoples R China
来源
16TH ANNUAL CONFERENCE OF THE INTERNATIONAL SPEECH COMMUNICATION ASSOCIATION (INTERSPEECH 2015), VOLS 1-5 | 2015年
关键词
rate of speech; deep neural network; speech recognition; SYLLABLE NUCLEI;
D O I
暂无
中图分类号
O42 [声学];
学科分类号
070206 ; 082403 ;
摘要
A significant performance reduction is often observed in speech recognition when the rate of speech (ROS) is too low or too high. Most of present approaches to addressing the ROS variation focus on the change of speech signals in dynamic properties caused by ROS, and accordingly modify the dynamic model, e.g., the transition probabilities of the hidden Markov model (HMM). However, an abnormal ROS changes not only the dynamic but also the static property of speech signals, and thus can not be compensated for purely by modifying the dynamic model. This paper proposes an ROS learning approach based on deep neural networks (DNN), which involves an ROS feature as the input of the DNN model and so the spectrum distortion caused by ROS can be learned and compensated for. The experimental results show that this approach can deliver better performance for too slow and too fast utterances, demonstrating our conjecture that ROS impacts both the dynamic and the static property of speech. In addition, the proposed approach can be combined with the conventional HMM transition adaptation method, offering additional performance gains.
引用
收藏
页码:528 / 532
页数:5
相关论文
共 17 条
  • [1] [Anonymous], 2014, DEEP LEARNING METHOD
  • [2] Chu S. M., 2010, ICASSP 10
  • [3] Praat script to detect syllable nuclei and measure speech rate automatically
    de Jong, Nivja H.
    Wempe, Ton
    [J]. BEHAVIOR RESEARCH METHODS, 2009, 41 (02) : 385 - 390
  • [4] Gupta V., 2014, ICASSP 14
  • [5] Heinrich C., 2011, P INT
  • [6] Li Y. hao, 2010, ISCSLP 10, P44
  • [7] Martinez F., 1998, ICASSP 98
  • [8] MORGAN N, 1997, P 5 EUR C SPEECH COM, V4, P2079
  • [9] Nikki Mirghafori, 1996, SPOK LANG ICSLP 96 P, V4
  • [10] Rouvier M., 2014, INT 14