|International Conference and Workshop on Emerging Trends in Technology
|Foundation of Computer Science USA
|ICWET - Number 1
|Authors: Dr. H. B. Kekre, Archana Athawale, Mrunali Desai
Dr. H. B. Kekre, Archana Athawale, Mrunali Desai . Text Dependent Speaker Identification using Row Mean Vector of Variable Sized Spectrogram. International Conference and Workshop on Emerging Trends in Technology. ICWET, 1 (None 2011), 43-47.
In this paper a simple approach to text dependent speaker identification using spectrograms and row mean is presented. This, mainly, revolves around trapping the complex patterns of variation in frequency and amplitude with time while an individual utters a given word through histogram equalized spectrogram. These histogram equalized spectrograms are used as a database to successfully identify the unknown individual from his/her voice. The features used for identifying, rely on optimal spectrogram segmentation and the Euclidean distance of the distributional features of the spectrograms of the unknown voice with that of a given known speaker in the database. Performance of this novel approach on a sample collected as two separate databases from 12 speakers and 28 speakers show that this methodology can be effectively used to produce a desirable success rate.