Hardware-based support vector machine for phoneme classification

Author

Cutajar, M. ; Gatt, E. ; Grech, I. ; Casha, O. ; Micallef, J.

Author_Institution

Dept. of Microelectron. & Nanoelectron., Univ. of Malta, Msida, Malta

fYear

2013

fDate

1-4 July 2013

Firstpage

1701

Lastpage

1708

Abstract

This paper presents the design of a digital hardware implementation based on Support Vector Machines (SVMs), for the task of multi-speaker phoneme recognition. The One-against-one multiclass SVM method, with the Radial Basis Function (RBF) kernel was considered. Furthermore, a priority scheme was also included in the architecture, in order to forecast the three most likely phonemes. The designed system was synthesised on a Xilinx Virtex-II XC2V3000 FPGA, and evaluated with the TIMIT corpus. This phoneme recognition system is intended to be implemented on a dedicated chip, along with the Discrete Wavelet Transforms (DWTs) for feature extraction, to further improve the resultant performance.

Keywords

discrete wavelet transforms; electronic design automation; feature extraction; field programmable gate arrays; radial basis function networks; speaker recognition; speech recognition equipment; support vector machines; DWT; FPGA; SVM method; TIMIT corpus; Xilinx Virtex-II XC2V3000; digital hardware implementation; discrete wavelet transforms; feature extraction; hardware based support vector machine; multispeaker phoneme recognition; one against one multiclass method; phoneme classification; phoneme recognition system; radial basis function; Accuracy; Computer architecture; Hardware; Kernel; Speech recognition; Support vector machines; field programmable gate arrays; phoneme recognition; speaker-independent; support vector machines;

fLanguage

English

Publisher

ieee

Conference_Titel

EUROCON, 2013 IEEE

Conference_Location

Zagreb

Print_ISBN

978-1-4673-2230-0

Type

conf

DOI

10.1109/EUROCON.2013.6625206

Filename

6625206