Speech Recognition using Processor DSP TMS320C6713 DSK

Abstract

Begin t. The speech signal transmits a lot of information on the speaker identity sex and emotion, numerous works have focused on the problem of automatic speaker recognition. The feature extraction and classification methods are the most important tasks in the speech recognition process.

As part of this work we are interested in the extraction of parameters by an algorithm that is inspired by the human auditory model, Mel Frequency Cepstral Coefficients (MFCC). The speech features extracted (MFCCs) are quantified using the algorithm of Vector Quantization.

In this paper we used TMS32006713 Kit with softwares Code Composer Studio and   Matlab for speech recognition in real time.