Abstract—Artificial intelligence advancements increased
speech usage as an interface for different processes. The idea of
Speaker based speech Recognition is basically to recognize
automatically who is speaking what; on the basis of the
individual information included in speech waves. That can only
be accomplished using speech recognition technique along with
speaker recognition algorithm. Among biometric recognition
techniques, speech recognition is one of the most useful tools.
Many banks, institutions, industries etc. are currently using
speaker based speech recognition for providing greater security
to their equipment. The idea is to implement this algorithm on
voice controlled robot incorporated with security features by
virtue of speech recognition algorithms. This robot will be
capable of recognizing ―What is being said and who has said it?‖
The algorithm is developed on MATLAB and implemented on a
TMS320C6713 DSP kit which will interface with a robot – A
voice controlled robot (VCR).
Index Terms—Mahalanobis distance, mel frequency cepstral
coefficients, speech recognition, vector quantization.
Maeira Imtiaz is with National University of Sciences and Technology,
Pakistan. She is also with MOL Pakistan Oil and Gas Co. B.V., Pakistan
(e-mail: maeiraimtiaz@yahoo.com).
Kamran A. Bhatti is with National University of Sciences and
Technology, Pakistan (e-mail: kambhatti2k@gmail.com).
[PDF]
Cite: Maeira Imtiaz and Kamran A. Bhatti, "Speech Dependent Speaker Classification Based Application Using DSPc6713," International Journal of Knowledge Engineering vol. 2, no. 1, pp. 31-34, 2016.
