Feature Extraction Using MFCC for Speech Recognition

Conference: Creative Trends in Engineering and Technology
Author(s): Sherry Vijh, Parminder Singh, Manjot Kaur Gill Year: 2016
Grenze ID: 02.CTET.2016.1.512 Page: 310-314


Speech is one of the stellar means of verbal expression used by humans to interact with the machines. Speech is\nproduced when air is rushed from lungs through vocal cords causing trembling of the vocal cords. With the introduction of\ndigital techniques and methods researchers could transform the speech signals in analog form to digital form at minimal cost\nand with much higher efficiency. Speech recognition is a mechanism to convert the digital speech signal into textual form.\nFeature extraction is the most significant phase of the mechanism which is concerned with extraction of features from given\nsignal. For efficacy of recognition mechanism robust features are required. In this paper focus is on mechanism of feature\nextraction from isolated words taking MFCC as the standard technique to extract the features. The paper highlights the\ncomparative advantage of MFCC and elaborates its complete procedure.


CTET - 2016