Cover Image

A Deep Learning Method Using Gender-Specific Features for Emotion Recognition

Speech reflects people’s mental state and using a microphone sensor is a potential method for human–computer interaction. Speech recognition using this sensor is conducive to the diagnosis of mental illnesses. The gender difference of speakers affects the process of speech emotion recognition based...

Full description

Saved in:
Bibliographic Details
Main Authors: Li-Min Zhang, Yang Li, Yue-Ting Zhang, Giap Weng Ng, Yu-Beng Leau, Hao Yan
Format: Article
Language:English
English
Published: Molecular Diversity Preservation International (MDPI) 2023
Subjects:
PN4775-4784 Technique. Practical journalism
QA75.5-76.95 Electronic computers. Computer science
Online Access:https://eprints.ums.edu.my/id/eprint/36091/1/ABSTRACT.pdf
https://eprints.ums.edu.my/id/eprint/36091/2/FULL%20TEXT.pdf
https://eprints.ums.edu.my/id/eprint/36091/
https://doi.org/10.3390/s23031355
Tags: Add Tag
No Tags, Be the first to tag this record!
id my.ums.eprints.36091
record_format eprints
spelling my.ums.eprints.360912023-07-21T06:51:54Z https://eprints.ums.edu.my/id/eprint/36091/ A Deep Learning Method Using Gender-Specific Features for Emotion Recognition Li-Min Zhang Yang Li Yue-Ting Zhang Giap Weng Ng Yu-Beng Leau Hao Yan PN4775-4784 Technique. Practical journalism QA75.5-76.95 Electronic computers. Computer science Speech reflects people’s mental state and using a microphone sensor is a potential method for human–computer interaction. Speech recognition using this sensor is conducive to the diagnosis of mental illnesses. The gender difference of speakers affects the process of speech emotion recognition based on specific acoustic features, resulting in the decline of emotion recognition accuracy. Therefore, we believe that the accuracy of speech emotion recognition can be effectively improved by selecting different features of speech for emotion recognition based on the speech representations of different genders. In this paper, we propose a speech emotion recognition method based on gender classification. First, we use MLP to classify the original speech by gender. Second, based on the different acoustic features of male and female speech, we analyze the influence weights of multiple speech emotion features in male and female speech, and establish the optimal feature sets for male and female emotion recognition, respectively. Finally, we train and test CNN and BiLSTM, respectively, by using the male and the female speech emotion feature sets. The results show that the proposed emotion recognition models have an advantage in terms of average recognition accuracy compared with gender-mixed recognition models Molecular Diversity Preservation International (MDPI) 2023 Article NonPeerReviewed text en https://eprints.ums.edu.my/id/eprint/36091/1/ABSTRACT.pdf text en https://eprints.ums.edu.my/id/eprint/36091/2/FULL%20TEXT.pdf Li-Min Zhang and Yang Li and Yue-Ting Zhang and Giap Weng Ng and Yu-Beng Leau and Hao Yan (2023) A Deep Learning Method Using Gender-Specific Features for Emotion Recognition. Sensors, 23. pp. 1-15. https://doi.org/10.3390/s23031355
institution Universiti Malaysia Sabah
building UMS Library
collection Institutional Repository
continent Asia
country Malaysia
content_provider Universiti Malaysia Sabah
content_source UMS Institutional Repository
url_provider http://eprints.ums.edu.my/
language English
English
topic PN4775-4784 Technique. Practical journalism
QA75.5-76.95 Electronic computers. Computer science
spellingShingle PN4775-4784 Technique. Practical journalism
QA75.5-76.95 Electronic computers. Computer science
Li-Min Zhang
Yang Li
Yue-Ting Zhang
Giap Weng Ng
Yu-Beng Leau
Hao Yan
A Deep Learning Method Using Gender-Specific Features for Emotion Recognition
description Speech reflects people’s mental state and using a microphone sensor is a potential method for human–computer interaction. Speech recognition using this sensor is conducive to the diagnosis of mental illnesses. The gender difference of speakers affects the process of speech emotion recognition based on specific acoustic features, resulting in the decline of emotion recognition accuracy. Therefore, we believe that the accuracy of speech emotion recognition can be effectively improved by selecting different features of speech for emotion recognition based on the speech representations of different genders. In this paper, we propose a speech emotion recognition method based on gender classification. First, we use MLP to classify the original speech by gender. Second, based on the different acoustic features of male and female speech, we analyze the influence weights of multiple speech emotion features in male and female speech, and establish the optimal feature sets for male and female emotion recognition, respectively. Finally, we train and test CNN and BiLSTM, respectively, by using the male and the female speech emotion feature sets. The results show that the proposed emotion recognition models have an advantage in terms of average recognition accuracy compared with gender-mixed recognition models
format Article
author Li-Min Zhang
Yang Li
Yue-Ting Zhang
Giap Weng Ng
Yu-Beng Leau
Hao Yan
author_facet Li-Min Zhang
Yang Li
Yue-Ting Zhang
Giap Weng Ng
Yu-Beng Leau
Hao Yan
author_sort Li-Min Zhang
title A Deep Learning Method Using Gender-Specific Features for Emotion Recognition
title_short A Deep Learning Method Using Gender-Specific Features for Emotion Recognition
title_full A Deep Learning Method Using Gender-Specific Features for Emotion Recognition
title_fullStr A Deep Learning Method Using Gender-Specific Features for Emotion Recognition
title_full_unstemmed A Deep Learning Method Using Gender-Specific Features for Emotion Recognition
title_sort deep learning method using gender-specific features for emotion recognition
publisher Molecular Diversity Preservation International (MDPI)
publishDate 2023
url https://eprints.ums.edu.my/id/eprint/36091/1/ABSTRACT.pdf
https://eprints.ums.edu.my/id/eprint/36091/2/FULL%20TEXT.pdf
https://eprints.ums.edu.my/id/eprint/36091/
https://doi.org/10.3390/s23031355
_version_ 1772812711185350656
score 13.211869

Similar Items