Study on the Behaviour of Mel Frequency Cepstral Coffecient Algorithm for Different Windows

Abstract

Mel Frequency Cepstral Coefficient or simply MFCC is a feature extracting algorithm can be applied on the real time signals. The Algorithm involves various steps and each step can be optimized mathematically, one of the stages is to apply a window to the signal for the signal processing proposes. There is list of windows which are actually can optimize the algorithm to get optimized. This paper notices each one of the window applications of the algorithm and its behaviours and based on the response of the windows to the signal input, particular segment of the algorithm can be modified. The modification of the small segment can lead us to the overall improvement of the MFCC algorithm.