Speech recognition on an FPGA using discrete and continuous hidden Markov models

Stephen Melnikoff, Steven Quigley, Martin Russell

Research output: Chapter in Book/Report/Conference proceedingConference contribution

11 Citations (Scopus)
269 Downloads (Pure)

Abstract

Speech recognition is a computationally demanding task, particularly the stage which uses Viterbi decoding for converting pre-processed speech data into words or sub-word units. Any device that can reduce the load on, for example, a PC’s processor, is advantageous. Hence we present FPGA implementations of the decoder based alternately on discrete and continuous hidden Markov models (HMMs) representing monophones, and demonstrate that the discrete version can process speech nearly 5,000 times real time, using just 12% of the slices of a Xilinx Virtex XCV1000, but with a lower recognition rate than the continuous implementation, which is 75 times faster than real time, and occupies 45% of the same device.
Original languageEnglish
Title of host publicationField-Programmable Logic and Applications. Reconfigurable Computing Is Going Mainstream 12th International Conference, FPL 2002, Montpellier, France September 2-4, 2002. Proceedings
PublisherSpringer
Pages202-211
Number of pages10
Publication statusPublished - 1 Jan 2002
Event12th International Conference on Field-Programmable Logic and Applications, Sep 02-04, 2002. MONTPELLIER, France -
Duration: 1 Jan 2002 → …

Publication series

NameLecture Notes in Computer Science
PublisherSpringer Verlag
Volume2438
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference12th International Conference on Field-Programmable Logic and Applications, Sep 02-04, 2002. MONTPELLIER, France
Period1/01/02 → …

Fingerprint

Dive into the research topics of 'Speech recognition on an FPGA using discrete and continuous hidden Markov models'. Together they form a unique fingerprint.

Cite this