The PDF file you selected should load here if your Web browser has a PDF reader plug-in installed (for example, a recent version of Adobe Acrobat Reader).

If you would like more information about how to print, save, and work with PDFs, Highwire Press provides a helpful Frequently Asked Questions about PDFs.

Alternatively, you can download the PDF file directly to your computer, from where it can be opened using a PDF reader. To download the PDF, click the Download link above.

Fullscreen Fullscreen Off


This paper presents the effectiveness of perceptual features and iterative clustering approach for performing both speech and speaker recognition. Procedure used for formation of training speech is different for developing training models for speaker independent speech and text independent speaker recognition. So, this work mainly emphasizes the utilization of clustering models developed for the training data to obtain better accuracy as 91%, 91% and 99.5% for mel frequency perceptual linear predictive cepstrum with respect to three categories such as speaker identification, isolated digit recognition and continuous speech recognition. This feature also produces 9% as low equal error rate which is used as a performance measure for speaker verification. The work is experimentally evaluated on the set of isolated digits and continuous speeches from TI digits_1 and TI digits_2 database for speech recognition and on speeches of 50 speakers randomly chosen from TIMIT database for speaker recognition. The noteworthy feature of speaker recognition algorithm is to evaluate the testing procedure on identical messages of all the 50 speakers, theoretical validation of results using F-ratio and validation of results by statistical analysis using χ2 distribution.

Keywords

Clustering Methods, Speech Recognition, Speaker Recognition, Spectral Analysis, Speech Analysis, Speech Processing, Vector Quantization.
User
Notifications
Font Size