Spectral Features Based on Local Hu Moments of Gabor Spectrograms for Speech Emotion Recognition

Huawei TAO; Ruiyu LIANG; Cheng ZHA; Xinran ZHANG; Li ZHAO

doi:10.1587/transinf.2015EDL8258

Spectral Features Based on Local Hu Moments of Gabor Spectrograms for Speech Emotion Recognition

Huawei TAO, Ruiyu LIANG, Cheng ZHA, Xinran ZHANG, Li ZHAO

Full Text Views

0

Cite this

Summary :

To improve the recognition rate of the speech emotion, new spectral features based on local Hu moments of Gabor spectrograms are proposed, denoted by GSLHu-PCA. Firstly, the logarithmic energy spectrum of the emotional speech is computed. Secondly, the Gabor spectrograms are obtained by convoluting logarithmic energy spectrum with Gabor wavelet. Thirdly, Gabor local Hu moments(GLHu) spectrograms are obtained through block Hu strategy, then discrete cosine transform (DCT) is used to eliminate correlation among components of GLHu spectrograms. Fourthly, statistical features are extracted from cepstral coefficients of GLHu spectrograms, then all the statistical features form a feature vector. Finally, principal component analysis (PCA) is used to reduce redundancy of features. The experimental results on EmoDB and ABC databases validate the effectiveness of GSLHu-PCA.

Publication: IEICE TRANSACTIONS on Information Vol.E99-D No.8 pp.2186-2189

Publication Date: 2016/08/01

Publicized: 2016/05/06

Online ISSN: 1745-1361

DOI: 10.1587/transinf.2015EDL8258

Type of Manuscript: LETTER

Category: Pattern Recognition

Authors

Huawei TAO
  Southeast University
Ruiyu LIANG
  Nanjing Institute of Technology
Cheng ZHA
  Southeast University
Xinran ZHANG
  Southeast University
Li ZHAO
  Southeast University

Keyword

Gabor wavelet, Hu invariant moment, speech emotion recognition, spectral features

Cite this

Copy

Huawei TAO, Ruiyu LIANG, Cheng ZHA, Xinran ZHANG, Li ZHAO, "Spectral Features Based on Local Hu Moments of Gabor Spectrograms for Speech Emotion Recognition" in IEICE TRANSACTIONS on Information, vol. E99-D, no. 8, pp. 2186-2189, August 2016, doi: 10.1587/transinf.2015EDL8258.
Abstract: To improve the recognition rate of the speech emotion, new spectral features based on local Hu moments of Gabor spectrograms are proposed, denoted by GSLHu-PCA. Firstly, the logarithmic energy spectrum of the emotional speech is computed. Secondly, the Gabor spectrograms are obtained by convoluting logarithmic energy spectrum with Gabor wavelet. Thirdly, Gabor local Hu moments(GLHu) spectrograms are obtained through block Hu strategy, then discrete cosine transform (DCT) is used to eliminate correlation among components of GLHu spectrograms. Fourthly, statistical features are extracted from cepstral coefficients of GLHu spectrograms, then all the statistical features form a feature vector. Finally, principal component analysis (PCA) is used to reduce redundancy of features. The experimental results on EmoDB and ABC databases validate the effectiveness of GSLHu-PCA.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.2015EDL8258/_p

Copy

@ARTICLE{e99-d_8_2186,
author={Huawei TAO, Ruiyu LIANG, Cheng ZHA, Xinran ZHANG, Li ZHAO, },
journal={IEICE TRANSACTIONS on Information},
title={Spectral Features Based on Local Hu Moments of Gabor Spectrograms for Speech Emotion Recognition},
year={2016},
volume={E99-D},
number={8},
pages={2186-2189},
abstract={To improve the recognition rate of the speech emotion, new spectral features based on local Hu moments of Gabor spectrograms are proposed, denoted by GSLHu-PCA. Firstly, the logarithmic energy spectrum of the emotional speech is computed. Secondly, the Gabor spectrograms are obtained by convoluting logarithmic energy spectrum with Gabor wavelet. Thirdly, Gabor local Hu moments(GLHu) spectrograms are obtained through block Hu strategy, then discrete cosine transform (DCT) is used to eliminate correlation among components of GLHu spectrograms. Fourthly, statistical features are extracted from cepstral coefficients of GLHu spectrograms, then all the statistical features form a feature vector. Finally, principal component analysis (PCA) is used to reduce redundancy of features. The experimental results on EmoDB and ABC databases validate the effectiveness of GSLHu-PCA.},
keywords={},
doi={10.1587/transinf.2015EDL8258},
ISSN={1745-1361},
month={August},}

Copy

TY - JOUR
TI - Spectral Features Based on Local Hu Moments of Gabor Spectrograms for Speech Emotion Recognition
T2 - IEICE TRANSACTIONS on Information
SP - 2186
EP - 2189
AU - Huawei TAO
AU - Ruiyu LIANG
AU - Cheng ZHA
AU - Xinran ZHANG
AU - Li ZHAO
PY - 2016
DO - 10.1587/transinf.2015EDL8258
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E99-D
IS - 8
JA - IEICE TRANSACTIONS on Information
Y1 - August 2016
AB - To improve the recognition rate of the speech emotion, new spectral features based on local Hu moments of Gabor spectrograms are proposed, denoted by GSLHu-PCA. Firstly, the logarithmic energy spectrum of the emotional speech is computed. Secondly, the Gabor spectrograms are obtained by convoluting logarithmic energy spectrum with Gabor wavelet. Thirdly, Gabor local Hu moments(GLHu) spectrograms are obtained through block Hu strategy, then discrete cosine transform (DCT) is used to eliminate correlation among components of GLHu spectrograms. Fourthly, statistical features are extracted from cepstral coefficients of GLHu spectrograms, then all the statistical features form a feature vector. Finally, principal component analysis (PCA) is used to reduce redundancy of features. The experimental results on EmoDB and ABC databases validate the effectiveness of GSLHu-PCA.
ER -