COUPLED UNSUPERVISED DEEP CONVOLUTIONAL DOMAIN ADAPTATION FOR SPEECH EMOTION RECOGNITION

The assumption of identical feature distributions in both training and testing datasets in Speech Emotion Recognition is not applicable in many scenarios, especially when different datasets are involved in the training and testing phases. Domain Adaptation methods have been proposed over the years to curtail this issue. In this paper, we propose an Unsupervised Domain Adaptation method which is a Coupled Deep Convolutional Neural Network (CDCNN) architecture. The architecture uses the correlation alignment loss (CALoss) of both the source and target distributions without target labels to effectively minimize the domain shift and learn extremely good nonlinear transformations. Also, the weights in the corresponding layers in both streams are not shared yet related which is effective for modeling the shift of one domain to the other. To evaluate our proposed method, we use the INTERSPEECH 2009 Emotion Challenge's FAU Aibo Emotion Corpus as target dataset and two publicly available corpora (ABC and Emo-DB) as source dataset. Experimental results indicate that our proposed method is superior to other state-of-the-arts-methods.