The objective of this research is to apply artificial neural network (ANN) to authenticate the disable users with their speech using MFCC features. To access the web content a speech authentication system is proposed for disable users.
We have used Genetic algorithm (GA) to create unique feature sets according to the objective function for the training of ANN for the acceptance of classification accuracy. This research work has dealt in the improvement of speech authentication system.
To extract the feature sets from the speech signals MFCC feature extraction technique is used with GA. In this work, GA is used as a feature optimization techniques and ANN as classifier to train the speech authentication system.
The evaluation and validation is being conducted on MATLAB 2016a simulator. To check the performance, parameters, like, precision, recall, f-measure, execution time and accuracy
have been calculated.
– Anand R Mehta – Download Document