Accent Recognition for Noisy Audio Signals

Ma, Zichen, and Fokoue, Ernest. "Accent Recognition for Noisy Audio Signals." Serdica Journal of Computing 8.2 (2014): 169-182. <http://eudml.org/doc/269895>.

@article{Ma2014,
abstract = {It is well established that accent recognition can be as accurate as up to 95% when the signals are noise-free, using feature extraction techniques such as mel-frequency cepstral coefficients and binary classifiers such as discriminant analysis, support vector machine and k-nearest neighbors. In this paper, we demonstrate that the predictive performance can be reduced by as much as 15% when the signals are noisy. Specifically, in this paper we perturb the signals with different levels of white noise, and as the noise become stronger, the out-of-sample predictive performance deteriorates from 95% to 80%, although the in-sample prediction gives overly-optimistic results. ACM Computing Classification System (1998): C.3, C.5.1, H.1.2, H.2.4., G.3.},
author = {Ma, Zichen, Fokoue, Ernest},
journal = {Serdica Journal of Computing},
keywords = {Ill-Posed Problem; Feature Extraction; Mel-Frequency Cepstral Coefficients; Discriminant Analysis; Support Vector Machine; K-Nearest Neighbors; Autoregressive Noise},
language = {eng},
number = {2},
pages = {169-182},
publisher = {Institute of Mathematics and Informatics Bulgarian Academy of Sciences},
title = {Accent Recognition for Noisy Audio Signals},
url = {http://eudml.org/doc/269895},
volume = {8},
year = {2014},
}

TY - JOUR
AU - Ma, Zichen
AU - Fokoue, Ernest
TI - Accent Recognition for Noisy Audio Signals
JO - Serdica Journal of Computing
PY - 2014
PB - Institute of Mathematics and Informatics Bulgarian Academy of Sciences
VL - 8
IS - 2
SP - 169
EP - 182
AB - It is well established that accent recognition can be as accurate as up to 95% when the signals are noise-free, using feature extraction techniques such as mel-frequency cepstral coefficients and binary classifiers such as discriminant analysis, support vector machine and k-nearest neighbors. In this paper, we demonstrate that the predictive performance can be reduced by as much as 15% when the signals are noisy. Specifically, in this paper we perturb the signals with different levels of white noise, and as the noise become stronger, the out-of-sample predictive performance deteriorates from 95% to 80%, although the in-sample prediction gives overly-optimistic results. ACM Computing Classification System (1998): C.3, C.5.1, H.1.2, H.2.4., G.3.
LA - eng
KW - Ill-Posed Problem; Feature Extraction; Mel-Frequency Cepstral Coefficients; Discriminant Analysis; Support Vector Machine; K-Nearest Neighbors; Autoregressive Noise
UR - http://eudml.org/doc/269895
ER -