gdg-speech-classifier

A machine learning system that recognizes the word 'Google' in human speech.

How it works

We train a classifier on a set of WAV files using Mel-Frequency Cepstral Coefficients (MFCC) as features. There are two implementations of the classifier available:

1. Regularized logistic regression, trained with conjugate gradient optimizer (fmincg).
2. Feed-forward neural network, trained with MATLAB's scaled conjugate gradient optimizer (trainscg).

How to use

1. Import training and test data into the data folder. You can get some data from the Releases Page. The names of the files should follow the pronunciation_en_%label%.wav pattern.
2. Run either mainLogisticRegression.m or mainNeuralNetwork.m depending on which classifier you would like to try.