Performance Enhancement of Phoneme Recognition using GPUs

DOI

10.2991/iccasp-16.2017.77How to use a DOI?

Keywords

Neural Networks, Back propagation, GPU.

Abstract

Phoneme recognition is a vital task in the automatic speech recognition process. The set of algorithms used for phoneme recognition requires effective handling of raw acoustic input data and demands huge amount of computations. New measures proposed in the pre-processing step and back propagation of artificial neural net-works attains improved accuracy and performance benefit, on its implementation on Graphics Processing Unit (GPU). The implementation has been tested on audio files with an average size of 250 KB. It is observed that there is a considerable performance enhancement of around 10x in the GPU implementation against CPU which portray considerable improvement in the learning accuracy.

Copyright

© 2017, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - P. Tamizharasan
AU  - M. Karthikeyan
AU  - N. Ramasubramanian
AU  - A. Joshi
PY  - 2016/12
DA  - 2016/12
TI  - Performance Enhancement of Phoneme Recognition using GPUs
BT  - Proceedings of the International Conference on Communication and Signal Processing 2016 (ICCASP 2016)
PB  - Atlantis Press
SP  - 524
EP  - 530
SN  - 1951-6851
UR  - https://doi.org/10.2991/iccasp-16.2017.77
DO  - 10.2991/iccasp-16.2017.77
ID  - Tamizharasan2016/12
ER  -