隐马尔可夫模型中一种新的帧相关建模方法
作者:
基金项目:

本文研究得到国家摼盼鍞攻关项目基金资助.


A New Method in Hidden Markov Model for Modeling Frame Correlation
  • 摘要
  • | |
  • 访问统计
  • |
  • 参考文献 [1]
  • |
  • 相似文献
  • |
  • 引证文献
  • | |
  • 文章评论
    摘要:

    在使用传统的隐马尔可夫模型(traditional hidden Markov model,简称THMM)刻画现实中的语音时有一个明显的缺点,即THMM不能合适地表征语音信号的时域结构.时域上的相关性被认为对识别非常有用,因为相邻帧间的特征矢量具有很强的相关性.文章提出了一种新的方法,用以把时域的相关性糅合到一个基于传统的隐马尔可夫模型的语音识别系统中.首先,用条件概率的形式处理帧间相关性;然后,用一种非线性的概率近似公式来表征相邻帧之间的相关性.此方法丝毫不增加原来的THMM的空间复杂度,而且也几乎不增

    Abstract:

    In this paper, the authors present a novel method to incorporate temporal correlation into a speech recognition system based on conventional hidden Markov model (HMM). The temporal correlation is considered to be useful for recognition because of the fact that the speech features of the present frame are highly informative about the feature characteristics of neighboring frames. An obvious way to incorporate temporal correlation is to condition the probability of the current observation on the current state as well as on the previous observation and the previous state. But using this method directly must lead to unreliable parameter estimation for the number of parameters to be estimated may increase too excessively to limited train data. In this paper, the authors approximate the joint conditional PD by non-linear estimation method. As a result, they can still use mixture Gaussian density to represent the joint conditional PD for the principle of any PD can be approximated by mixture Gaussian density. The HMM incorporated temporal correlation by non-linear estimation method, which they called FC (frame correlation) HMM does not need any additional parameters and it only brings a little additional computing quantity. The results of the experiment show that the top 1 recognition rate of FC HMM has been raised by 6 percent compared to the conventional HMM method.

    参考文献
    1  Ostendorf M, Roukos S. A stochastic segment model for phoneme-based continuous speech recognition. IEEE Transactions on Acoustics, Speech and Signal Processing, 1989,37(12):1857~1869 2  Digalakis V, Rohlicek J R, Ostendorf M. A dynamical system approach to continuous speech recognition. In: Proceedings of the International Conference Acoustics, Speech, and Signal Processing. Mississauga: Imperial Press Limited, 1991. 289~292 3  Wellekens C J. Explicit correlation in hidden Markov model for speech recognition. In: Proceedings of the International Conference Acoustics, Speech, and Signal Processing. San Francisco: IEEE Signal Processing Society, 1987. 383~386 4  Kenny P, Lennig M, Mermelstein P. A linear predictive HMM for vector-valued observations with applications to speech recognition. IEEE Transactions on Acoustics, Speech and Signal Processing, 1990,38(2):220~225 5  Paliwal K K. Use of temporal correlation between successive frames in hidden Markov model based speech recognizer. In: Proceedings of the International Conference Acoustics, Speech, and Signal Processing. San Francisco: IEEE Signal Processing Society, 1993. 215~218 6  Takahashi S. Phonemic HMM constrained by statistical VQ-code transition. In: Proceedings of the International Conference Acoustics, Speech, and Signal Processing. San Francisco: IEEE Signal Processing Society, 1992. 553~556 7  Takahashi S. Phoneme HMM's constrained by frame correlation. In: Proceedings of the International Conference Acoustics, Speech, and Signal Processing. San Francisco: IEEE Signal Processing Society, 1993. 219~222 8  Nam Soo Kim, Chong Kwan Un. Frame-correlated hidden Markov model based on extended logarithmic pool. IEEE Transactions on Speech and Audio Processing, 1997,5(2):149~160
    相似文献
    网友评论
    网友评论
    分享到微博
    发 布
引用本文

郭 庆,吴文虎,方棣棠.隐马尔可夫模型中一种新的帧相关建模方法.软件学报,1999,10(6):631-635

复制
分享
文章指标
  • 点击次数:3771
  • 下载次数: 4919
  • HTML阅读次数: 0
  • 引用次数: 0
历史
  • 收稿日期:1998-04-27
  • 最后修改日期:1998-06-23
文章二维码
您是第19794288位访问者
版权所有:中国科学院软件研究所 京ICP备05046678号-3
地址:北京市海淀区中关村南四街4号,邮政编码:100190
电话:010-62562563 传真:010-62562533 Email:jos@iscas.ac.cn
技术支持:北京勤云科技发展有限公司

京公网安备 11040202500063号