Skip to content

Publications

Detail of publication

Citation

Vaněk J. and Trmal J. and Psutka J. V. and Psutka J. : Full Covariance Gaussian Mixture Models Evaluation on GPU . IEEE International Symposium on Signal Processing and Information Technology, Vietnam, Ho Chi Minh City, 2012.

Download PDF

PDF

Abstract

Gaussian mixture models (GMMs) are often used in various data processing and classification tasks to model a continuous probability density in a multi-dimensional space. In cases, where the dimension of the feature space is relatively high (e.g. in the automatic speech recognition (ASR)), GMM with a higher number of Gaussians with diagonal covariances (DC) instead of full covariances (FC) is used from the two reasons. The first reason is a~problem how to estimate robust FC matrices with a~limited training data set. The second reason is a~much higher computational cost during the GMM evaluation. The first reason was addressed in many recent publications. In contrast, this paper describes an efficient implementation on Graphic Processing Unit (GPU) of the FC-GMM evaluation, which addresses the second reason. The performance was tested on acoustic models for ASR, and it is shown that even a low-end laptop GPU is capable to evaluate a large acoustic model in a fraction of the real speech time. Three variants of the algorithm were implemented and compared on various GPUs: NVIDIA CUDA, NVIDIA OpenCL, and ATI/AMD OpenCL.

Detail of publication

Title: Full Covariance Gaussian Mixture Models Evaluation on GPU
Author: Vaněk J. ; Trmal J. ; Psutka J. V. ; Psutka J.
Language: English
Date of publication: 12 Dec 2012
Year: 2012
Type of publication: Papers in proceedings of reviewed conferences
Title of journal or book: IEEE International Symposium on Signal Processing and Information Technology
ISBN: 978-1-4673-5604-6
Address: Vietnam, Ho Chi Minh City
Date: 12 Dec 2012 - 15 Dec 2012
/ 2014-11-12 12:23:33 /

Keywords

Gaussian Mixture Models, Full Covariance, Automatic Speech Recognition, GPU, CUDA, OpenCL

BibTeX

@INPROCEEDINGS{VanekJ_2012_FullCovariance,
 author = {Van\v{e}k J. and Trmal J. and Psutka J. V. and Psutka J.},
 title = {Full Covariance Gaussian Mixture Models Evaluation on GPU},
 year = {2012},
 journal = {IEEE International Symposium on Signal Processing and Information Technology},
 address = {Vietnam, Ho Chi Minh City},
 ISBN = {978-1-4673-5604-6},
 url = {http://www.kky.zcu.cz/en/publications/VanekJ_2012_FullCovariance},
}