GMM-HMM语音识别模型

• GMM-HMM语音识别模型 原理篇
  http://blog.csdn.net/abcjennifer/article/details/27346787
• 声学模型
  https://zh.wikipedia.org/wiki/%E5%A3%B0%E5%AD%A6%E6%A8%A1%E5%9E%8B
• Hidden Markov Models and Gaussian Mixture Models
  https://www.inf.ed.ac.uk/teaching/courses/asr/2012-13/asr03-hmmgmm-4up.pdf
• GMM-HMM ACOUSTIC MODEL TRAINING BY A TWO LEVEL PROCEDURE WITH GAUSSIAN COMPONENTS DETERMINED BY AUTOMATIC MODEL SELECTION
  http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=5495122
• Gaussian Mixture Model and Hidden Markov Model
  http://speech.tifr.res.in/tutorialSlides/gmmHmmTutoChiefWiSSAP09.pdf
• 语音识别系统原理介绍—从gmm-hmm到dnn-hmm
  http://blog.csdn.net/wbgxx333/article/details/18516053
• http://www.cs.ubbcluj.ro/~csatol/mach_learn/bemutato/PuskasJanos_GMMforSpeech.pdf
• http://www.cs.tut.fi/~puhtunn/lecture_model_search.pdf
• 《IMPROVED HYBRID MODEL OF HMM/GMM FOR SPEECH RECOGNITION》
  http://foibg.com/ibs_isc/ibs-05/IBS-05-p11.pdf
• 《Hidden Markov model/Gaussian mixture models(HMM/GMM) based voice command system: A way to improve the control of remotely operated robot arm》
  http://www.academicjournals.org/app/webroot/article/article1380631495_El-emary%20et%20al.pdf
• 《Speaker Verification Using Adapted Gaussian Mixture Models》
  http://web.cs.swarthmore.edu/~turnbull/cs97/f09/paper/reynolds00.pdf
• 语音识别的技术原理是什么
  http://www.zhihu.com/question/20398418
• 输出值是从各个帧计算而得的声学特征，特征是多为向量，特征取值可以是离散或连读的。
• 最下面的observation就是我们提取的特征。
  gmm-hmm就是把我们的特征用混合高斯模型区模拟，然后把均值和方差输入到hmm的模型里。
• 
• GMM-HMM语音识别简单理解 http://www.cnblogs.com/tornadomeet/p/3276753.html
• BLSTM-DNN hybrid语音识别声学模型的第一个工业应用 http://yq.aliyun.com/articles/2308?&utm_campaign=sys&utm_medium=market&utm_source=edm_email&msctype=email&mscmsgid=109016010600555289&