A Deep Learning Method Using Gender-Specific Features for Emotion Recognition
Sensors · 2023
DOI: 10.3390/s23031355 ↗Kurz gefasst
Fachartikel in der Zeitschrift Sensors, 2023. Verfasst von Li-Min Zhang und 5 weiteren Autor:innen. Beteiligte Einrichtung: Key Laboratory for Artificial Intelligence and Cognitive Neuroscience of Language, Xi'an International Studies University, Xi'an 610116, China. Hauptthemen laut MEDLINE: Deep Learning. Die Arbeit wurde bisher 17-mal zitiert (Crossref) und nennt 36 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Social Science Foundation of Shaanxi Province of China, National Social Science Foundation of China, Natural Science Basic Research Program of Shaanxi Province of China.
Abstract (Originalsprache)
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.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Key Laboratory for Artificial Intelligence and Cognitive Neuroscience of Language, Xi'an International Studies University, Xi'an 610116, China.
- Faculty of Computing and Informatics, Universiti Malaysia Sabah, Sabah 88400, Malaysia.
ORCID-Profile: Yu-Beng Leau
Förderung
- Social Science Foundation of Shaanxi Province of China
- National Social Science Foundation of China
- Natural Science Basic Research Program of Shaanxi Province of China
- Shaanxi Educational Science and Planning Foundation for “14th Five-Year Plan” of China
Zitieren
APA
Zhang, L. M., Li, Y., Zhang, Y. T., Ng, G. W., Leau, Y. B., & Yan, H. (2023). A Deep Learning Method Using Gender-Specific Features for Emotion Recognition. Sensors. https://doi.org/10.3390/s23031355
Vancouver
Zhang LM, Li Y, Zhang YT, Ng GW, Leau YB, Yan H. A Deep Learning Method Using Gender-Specific Features for Emotion Recognition. Sensors. 2023. doi:10.3390/s23031355
BibTeX
@article{Zhang2023,
author = {Zhang, Li-Min and Li, Yang and Zhang, Yue-Ting and Ng, Giap Weng and Leau, Yu-Beng and Yan, Hao},
title = {A Deep Learning Method Using Gender-Specific Features for Emotion Recognition},
journal = {Sensors},
year = {2023},
doi = {10.3390/s23031355}
}
EUVIMED