Towards real-time speech emotion recognition using deep neural networks

Fayek, H, Lech, M and Cavedon, L 2015, 'Towards real-time speech emotion recognition using deep neural networks', in Tadeusz A Wysocki and Beata J Wysocki (ed.) Proceedings of the 9th International Conference on Signal Processing and Communication Systems (ICSPCS 2015), Cairns, Australia, 14-16 December 2015, pp. 1-5.


Document type: Conference Paper
Collection: Conference Papers

Title Towards real-time speech emotion recognition using deep neural networks
Author(s) Fayek, H
Lech, M
Cavedon, L
Year 2015
Conference name ICSPCS 2015
Conference location Cairns, Australia
Conference dates 14-16 December 2015
Proceedings title Proceedings of the 9th International Conference on Signal Processing and Communication Systems (ICSPCS 2015)
Editor(s) Tadeusz A Wysocki and Beata J Wysocki
Publisher IEEE
Place of publication United States
Start page 1
End page 5
Total pages 5
Abstract Most existing Speech Emotion Recognition (SER) systems rely on turn-wise processing, which aims at recognizing emotions from complete utterances and an overly-complicated pipeline marred by many preprocessing steps and hand-engineered features. To overcome both drawbacks, we propose a real-time SER system based on end-to-end deep learning. Namely, a Deep Neural Network (DNN) that recognizes emotions from a one second frame of raw speech spectrograms is presented and investigated. This is achievable due to a deep hierarchical architecture, data augmentation, and sensible regularization. Promising results are reported on two databases which are the eNTERFACE database and the Surrey Audio-Visual Expressed Emotion (SAVEE) database.
Subjects Signal Processing
Keyword(s) affective computing
deep neural networks
deep learning
emotion recognition
spectrograms
speech processing
Copyright notice © 2015 IEEE
ISBN 9781467381185
Versions
Version Filter Type
Citation counts: TR Web of Science Citation Count  Cited 0 times in Thomson Reuters Web of Science Article
Scopus Citation Count Cited 10 times in Scopus Article | Citations
Access Statistics: 158 Abstract Views  -  Detailed Statistics
Created: Fri, 11 Mar 2016, 10:17:00 EST by Catalyst Administrator
© 2014 RMIT Research Repository • Powered by Fez SoftwareContact us