Evaluating deep learning architectures for Speech Emotion Recognition. (August 2017)

• Record Type: Journal Article
• Authors: Fayek, Haytham M.; Lech, Margaret; Cavedon, Lawrence
• Is Part Of: Neural networks. Volume 92(2017)
• Page Start: 60
• Page End: 68
• Publication Date: 2017-08
• Subjects: Affective computing -- Deep learning -- Emotion recognition -- Neural networks -- Speech recognition; Neural computers -- Periodicals; Neural networks (Computer science) -- Periodicals; Neural networks (Neurobiology) -- Periodicals; Nervous System -- Periodicals; Ordinateurs neuronaux -- Périodiques; Réseaux neuronaux (Informatique) -- Périodiques; Réseaux neuronaux (Neurobiologie) -- Périodiques; Neural computers; Neural networks (Computer science); Neural networks (Neurobiology); Periodicals; 006.32
• DOI: 10.1016/j.neunet.2017.02.013
• Languages: English
• ISSNs: 0893-6080

Abstract

Abstract: Speech Emotion Recognition (SER) can be regarded as a static or dynamic classification problem, which makes SER an excellent test bed for investigating and comparing various deep learning architectures. We describe a frame-based formulation to SER that relies on minimal speech processing and end-to-end deep learning to model intra-utterance dynamics. We use the proposed SER system to empirically explore feed-forward and recurrent neural network architectures and their variants. Experiments conducted illuminate the advantages and limitations of these architectures in paralinguistic speech recognition and emotion recognition in particular. As a result of our exploration, we report state-of-the-art results on the IEMOCAP database for speaker-independent SER and present quantitative and qualitative assessments of the models' performances.