Title :
Multicore processor focused on voice biometrics
Author :
Chovancova, E. ; Dudlakova, Z. ; Fortotira, O. ; Radusovsky, J.
Author_Institution :
Dept. of Comput. & Inf., Tech. Univ. of Kosice, Kosice, Slovakia
Abstract :
Biometrics deals with technology of measuring and analyzing biological data. In informatics, biometry refers to technologies that measure and analyze human body characteristics, such as a DNA, a fingerprint, a retina or iris, a voice patterns, a facial patterns and hand, for authentication purposes. This work points out the suggestion of a specialized processor focused on sound biometrics using Dynamic Time Warping method. This method is based on an analysis of the issue for most effective solution in VHDL. This method deals with recognition of certain words in the sentences, which are compared with stored templates. Therefor is it in terms of programming most appropriate and reasonable for processor design.
Keywords :
biometrics (access control); coprocessors; hardware description languages; speech recognition; word processing; VHDL; biological data analysis; dynamic time warping method; multicore processor design; sound biometrics; voice biometrics; word recognition; Hidden Markov models; Iris recognition; Random access memory; Registers; Speech recognition;
Conference_Titel :
Emerging eLearning Technologies and Applications (ICETA), 2014 IEEE 12th International Conference on
Print_ISBN :
978-1-4799-7739-0
DOI :
10.1109/ICETA.2014.7107551