Continuous Bangla Speech Segmentation using Short-term Speech Features Extraction Approaches
M. Md Mijanur Rahman. International Journal of Advanced Computer Science and Applications(IJACSA), (2012)
Abstract
This paper presents simple and novel feature extraction approaches for segmenting continuous Bangla speech sentences into words/sub-words. These methods are based on two simple speech features, namely the time-domain features and the frequency-domain features. The time-domain features, such as short-time signal energy, short-time average zero crossing rate and the frequency-domain features, such as spectral centroid and spectral flux features are extracted in this research work. After the feature sequences are extracted, a simple dynamic thresholding criterion is applied in order to detect the word boundaries and label the entire speech sentence into a sequence of words/sub-words. All the algorithms used in this research are implemented in Matlab and the implemented automatic speech segmentation system achieved segmentation accuracy of 96\%.
%0 Journal Article
%1 IJACSA.2012.031121
%A Md Mijanur Rahman, Md. Al-Amin Bhuiyan
%D 2012
%J International Journal of Advanced Computer Science and Applications(IJACSA)
%K Centroid; Dynamic Energy; Extraction; Features Segmentation; Short-time Spectral Speech Thresholding.
%N 11
%T Continuous Bangla Speech Segmentation using Short-term Speech Features Extraction Approaches
%U http://ijacsa.thesai.org/
%V 3
%X This paper presents simple and novel feature extraction approaches for segmenting continuous Bangla speech sentences into words/sub-words. These methods are based on two simple speech features, namely the time-domain features and the frequency-domain features. The time-domain features, such as short-time signal energy, short-time average zero crossing rate and the frequency-domain features, such as spectral centroid and spectral flux features are extracted in this research work. After the feature sequences are extracted, a simple dynamic thresholding criterion is applied in order to detect the word boundaries and label the entire speech sentence into a sequence of words/sub-words. All the algorithms used in this research are implemented in Matlab and the implemented automatic speech segmentation system achieved segmentation accuracy of 96\%.
@article{IJACSA.2012.031121,
abstract = {This paper presents simple and novel feature extraction approaches for segmenting continuous Bangla speech sentences into words/sub-words. These methods are based on two simple speech features, namely the time-domain features and the frequency-domain features. The time-domain features, such as short-time signal energy, short-time average zero crossing rate and the frequency-domain features, such as spectral centroid and spectral flux features are extracted in this research work. After the feature sequences are extracted, a simple dynamic thresholding criterion is applied in order to detect the word boundaries and label the entire speech sentence into a sequence of words/sub-words. All the algorithms used in this research are implemented in Matlab and the implemented automatic speech segmentation system achieved segmentation accuracy of 96\%.},
added-at = {2014-02-21T08:00:08.000+0100},
author = {{Md Mijanur Rahman}, Md. Al-Amin Bhuiyan},
biburl = {https://www.bibsonomy.org/bibtex/2b8f002ce43709e6528fa293062eb5890/thesaiorg},
interhash = {3d7ba77b05428d359c5fc52e32722bff},
intrahash = {b8f002ce43709e6528fa293062eb5890},
journal = {International Journal of Advanced Computer Science and Applications(IJACSA)},
keywords = {Centroid; Dynamic Energy; Extraction; Features Segmentation; Short-time Spectral Speech Thresholding.},
number = 11,
timestamp = {2014-02-21T08:00:08.000+0100},
title = {{Continuous Bangla Speech Segmentation using Short-term Speech Features Extraction Approaches}},
url = {http://ijacsa.thesai.org/},
volume = 3,
year = 2012
}