As a kind of statistical method, the technique of Hidden Markov Model （HMM） is widely used for speech recognition. In order to train the HMM to be more effective with much less amount of data, the Subspace Distribut...As a kind of statistical method, the technique of Hidden Markov Model （HMM） is widely used for speech recognition. In order to train the HMM to be more effective with much less amount of data, the Subspace Distribution Clustering Hidden Markov Model （SDCHMM）, derived from the Continuous Density Hidden Markov Model （CDHMM）, is introduced. With parameter tying, a new method to train SDCHMMs is described. Compared with the conventional training method, an SDCHMM recognizer trained by means of the new method achieves higher accuracy and speed. Experiment results show that the SDCHMM recognizer outperforms the CDHMM recognizer on speech recognition of Chinese digits.展开更多
An extended phonemic set of mandarin from the view of speech recognition is proposed. This set absorbs most principles of some other existing phonemic sets for mandarin, like Worldbet and SAMPA-C, and also takes advan...An extended phonemic set of mandarin from the view of speech recognition is proposed. This set absorbs most principles of some other existing phonemic sets for mandarin, like Worldbet and SAMPA-C, and also takes advantage of some practical experiences from speech recognition research for increasing the discriminability between word models. And the experiments in speaker independent continuous speech recognition show that hidden Markov models defined by this phonemic set have a better performance than those based on initial/final units of mandarin and have a very compact size.展开更多
The use of hidden conditional random fields （HCRFs） for tone modeling is explored. The tone recognition performance is improved using HCRFs by taking advantage of intra-syllable dynamic, inter-syllable dynamic and d...The use of hidden conditional random fields （HCRFs） for tone modeling is explored. The tone recognition performance is improved using HCRFs by taking advantage of intra-syllable dynamic, inter-syllable dynamic and duration features. When the tone model is integrated into continuous speech recognition, the discriminative model weight training （DMWT） is proposed. Acoustic and tone scores are scaled by model weights discriminatively trained by the minimum phone error （MPE） criterion. Two schemes of weight training are evaluated and a smoothing technique is used to make training robust to overtraining problem. Experiments show that the accuracies of tone recognition and large vocabulary continuous speech recognition （LVCSR） can be improved by the HCRFs based tone model. Compared with the global weight scheme, continuous speech recognition can be improved by the discriminative trained weight combinations.展开更多
A noise estimator was presented in this paper by modeling the log-power sequence with hidden Markov model （HMM）. The smoothing factor of this estimator was motivated by the speech presence probability at each freque...A noise estimator was presented in this paper by modeling the log-power sequence with hidden Markov model （HMM）. The smoothing factor of this estimator was motivated by the speech presence probability at each frequency band. This HMM had a speech state and a nonspeech state, and each state consisted of a unique Gaussian function. The mean of the nonspeech state was the estimation of the noise logarithmic power. To make this estimator run in an on-line manner, an HMM parameter updated method was used based on a first-order recursive process. The noise signal was tracked together with the HMM to be sequentially updated. For the sake of reliability, some constraints were introduced to the HMM. The proposed algorithm was compared with the conventional ones such as minimum statistics （MS） and improved minima controlled recursive averaging （IM- CRA）. The experimental results confirms its promising performance.展开更多
为了进行连续马尔可夫模型的初值提取,提出一种各类在训练样本空间近似均衡分布的K均值聚类法。在聚类的过程中引入惩罚因子,从而限制过多的训练矢量集中于一个或几个类,使样本空间划分近似均匀。连续马尔可夫模型初值提取实验证明,该...为了进行连续马尔可夫模型的初值提取,提出一种各类在训练样本空间近似均衡分布的K均值聚类法。在聚类的过程中引入惩罚因子,从而限制过多的训练矢量集中于一个或几个类,使样本空间划分近似均匀。连续马尔可夫模型初值提取实验证明,该方法与标准的K均值聚类法、LBG(L inde Buzo G ray)聚类法相比,降低了矢量量化产生的全局失真,各个类在样本空间的分布更加均匀,提高了矢量量化的性能。将该方法用于孤立词识别连续马尔可夫模型的初值提取,可使各个高斯概率密度函数的参数估计更逼近其无偏估计,从而提高了马尔可夫模型初值的可靠性。展开更多
基金Supported by the National Natural Science Foundation of China (No.60172048)
文摘As a kind of statistical method, the technique of Hidden Markov Model （HMM） is widely used for speech recognition. In order to train the HMM to be more effective with much less amount of data, the Subspace Distribution Clustering Hidden Markov Model （SDCHMM）, derived from the Continuous Density Hidden Markov Model （CDHMM）, is introduced. With parameter tying, a new method to train SDCHMMs is described. Compared with the conventional training method, an SDCHMM recognizer trained by means of the new method achieves higher accuracy and speed. Experiment results show that the SDCHMM recognizer outperforms the CDHMM recognizer on speech recognition of Chinese digits.
文摘An extended phonemic set of mandarin from the view of speech recognition is proposed. This set absorbs most principles of some other existing phonemic sets for mandarin, like Worldbet and SAMPA-C, and also takes advantage of some practical experiences from speech recognition research for increasing the discriminability between word models. And the experiments in speaker independent continuous speech recognition show that hidden Markov models defined by this phonemic set have a better performance than those based on initial/final units of mandarin and have a very compact size.
文摘The use of hidden conditional random fields （HCRFs） for tone modeling is explored. The tone recognition performance is improved using HCRFs by taking advantage of intra-syllable dynamic, inter-syllable dynamic and duration features. When the tone model is integrated into continuous speech recognition, the discriminative model weight training （DMWT） is proposed. Acoustic and tone scores are scaled by model weights discriminatively trained by the minimum phone error （MPE） criterion. Two schemes of weight training are evaluated and a smoothing technique is used to make training robust to overtraining problem. Experiments show that the accuracies of tone recognition and large vocabulary continuous speech recognition （LVCSR） can be improved by the HCRFs based tone model. Compared with the global weight scheme, continuous speech recognition can be improved by the discriminative trained weight combinations.
基金Supported by the National Key Basic Research Program of China(2013CB329302)the National Natural Science Foundation of China(61271426,10925419,90920302,61072124,11074275,11161140319,91120001)+3 种基金the Strategic Priority Research Program of the Chinese Academy of Sciences(XDA06030100,XDA06030500)the National "863" Program(2012AA012503)the CAS Priority Deployment Project(KGZD-EW-103-2)Jiangxi Provincial Department of Education Science and Technology Project(GJJ13426)
文摘A noise estimator was presented in this paper by modeling the log-power sequence with hidden Markov model （HMM）. The smoothing factor of this estimator was motivated by the speech presence probability at each frequency band. This HMM had a speech state and a nonspeech state, and each state consisted of a unique Gaussian function. The mean of the nonspeech state was the estimation of the noise logarithmic power. To make this estimator run in an on-line manner, an HMM parameter updated method was used based on a first-order recursive process. The noise signal was tracked together with the HMM to be sequentially updated. For the sake of reliability, some constraints were introduced to the HMM. The proposed algorithm was compared with the conventional ones such as minimum statistics （MS） and improved minima controlled recursive averaging （IM- CRA）. The experimental results confirms its promising performance.
文摘为了进行连续马尔可夫模型的初值提取,提出一种各类在训练样本空间近似均衡分布的K均值聚类法。在聚类的过程中引入惩罚因子,从而限制过多的训练矢量集中于一个或几个类,使样本空间划分近似均匀。连续马尔可夫模型初值提取实验证明,该方法与标准的K均值聚类法、LBG(L inde Buzo G ray)聚类法相比,降低了矢量量化产生的全局失真,各个类在样本空间的分布更加均匀,提高了矢量量化的性能。将该方法用于孤立词识别连续马尔可夫模型的初值提取,可使各个高斯概率密度函数的参数估计更逼近其无偏估计,从而提高了马尔可夫模型初值的可靠性。
