Speech recognition using dynamic time warping DTW in Matlab

This code analyzes user voice saying digits and train the computer to it so it can be identified later and determine the digits said by the user. This code can be a core for very exciting projects like turning speech to text or making a voice password that can't be hacked.

The code is implemented in four scripts two for creating the training set and another for creating input voice files. Another two for extracting the voice features using MFCC algorithm. Then the features are saved in Features.mat file and the final code takes in the voice input and recognize the digits said by user by comparing the input voice to the saved one using dynamic time warping DTW algorithm. The code works with high accuracy on MATLAB platform.

In time series analysis, dynamic time warping (DTW) is one of the algorithms for measuring similarity between two temporal sequences, which may vary in speed. In general, DTW is a method that calculates an optimal match between two given sequences (e.g. sound signals).

Mel-frequency cepstral coefficients (MFCCs) are coefficients that collectively make up an MFC. They are derived from a type of cepstral representation of the audio clip and used as unique features for the audio signal

Contact me:

email: s-ahmed.saleh@zewailcity.edu.eg

List of all my codes: https://gumroad.com/ahmedsaleh

===================================================

Hire me directly on freelancer:

https://www.freelancer.com/u/AhmedSobhiSaleh

===================================================

DTW, dynamic time warping, MFCC, speech recognition,voice recognition,digits,speech to text, machine learning, voice orders, voice analysis, audio signal, signal processing, training set, testing set,