Java Speech Recognition using Hidden Markov Model / Vector Quantization / MFCC

Automatically exported from code.google.com/p/speech-recognition-java-hidden-markov-model-vq-mfcc

===

Accompanying project report

• http://ganeshtiwaridotcomdotnp.blogspot.com/2011/06/final-report-text-prompted-remote.html

Main classes to look into are :

• org.ioe.tprsa.mediator.Operations : class that demonstrates use of HMM, VQ training and testing

• org.ioe.tprsa.ui.HMM_VQ_Speech_Recognition : GUI that allows record of voice samples per word, train the samples, test with a recorded audio or saved .wav file.

Folder conventions :

Training audio files

• speechTrainWav\apple\apple01.wav
• speechTrainWav\apple\apple02.wav
• speechTrainWav\apple\apple03.wav
• speechTrainWav\apple\apple04.wav
• speechTrainWav\cat\cat01.wav
• speechTrainWav\cat\cat02.wav
• speechTrainWav\cat\cat03.wav

Vector Quantization code book :

• models\codeBook\codebook.cbk

HMM trained models :

• models\HMM\apple.hmm
• models\HMM\cat.hmm
• models\GMM\Nepal.hmm

Please feel free to use/modify the code !

For any queries :

Email : gtiwari333@gmail.com

Blog : http://ganeshtiwaridotcomdotnp.blogspot.com/