Showing 1–1 of 1 results for author: Kadoch, D
-
Towards Robust Deep Neural Networks for Affect and Depression Recognition from Speech
Authors:
Alice Othmani,
Daoud Kadoch,
Kamil Bentounes,
Emna Rejaibi,
Romain Alfred,
Abdenour Hadid
Abstract:
Intelligent monitoring systems and affective computing applications have emerged in recent years to enhance healthcare. Examples of these applications include assessment of affective states such as Major Depressive Disorder (MDD). MDD describes the constant expression of certain emotions: negative emotions (low Valence) and lack of interest (low Arousal). High-performing intelligent systems would…
▽ More
Intelligent monitoring systems and affective computing applications have emerged in recent years to enhance healthcare. Examples of these applications include assessment of affective states such as Major Depressive Disorder (MDD). MDD describes the constant expression of certain emotions: negative emotions (low Valence) and lack of interest (low Arousal). High-performing intelligent systems would enhance MDD diagnosis in its early stages. In this paper, we present a new deep neural network architecture, called EmoAudioNet, for emotion and depression recognition from speech. Deep EmoAudioNet learns from the time-frequency representation of the audio signal and the visual representation of its spectrum of frequencies. Our model shows very promising results in predicting affect and depression. It works similarly or outperforms the state-of-the-art methods according to several evaluation metrics on RECOLA and on DAIC-WOZ datasets in predicting arousal, valence, and depression. Code of EmoAudioNet is publicly available on GitHub: https://github.com/AliceOTHMANI/EmoAudioNet
△ Less
Submitted 18 November, 2020; v1 submitted 1 November, 2019;
originally announced November 2019.
