Singer identification in Indian Hindi songs using MFCC and spectral features

In this paper, there is an attempt to give a simple and efficient solution to the problem of singer identification in Indian Hindi songs using audio samples. The experiment is done using five well known singers. This work uses MFCC and some spectral features which are extracted for each audio sample taken of size 2 seconds. After a set of pre-processing steps, supervised training is done using multilayer feed-forward artificial neural network and back propagation algorithm. The results obtained strongly suggest that the final model is able to perform the task of singer identification in Indian Hindi songs with good accuracy.