Comparison between weighted D-KNN and other classifiers for mandarin speech emotion recognition

The interaction between human beings and computers will be more natural if computers are able to perceive and respond to human non-verbal communication such as emotion. Whereas research about automated recognition of emotions in facial expressions is now very rich, research dealing with the speech modality has only been active for recent years and adopts several classifiers for automatically assigning an emotion category, e.g., anger, happiness or sadness, to a speech utterance. These classifiers were designed independently and tested on separate emotional speech corpora, making it difficult to compare and evaluate their performance. In this paper, we compared the performance of the proposed weighted D-KNN classification method with that of other popular classifiers by applying them to a Mandarin speech corpus consisting of five basic emotions. The experimental results show that all classification techniques have their own advantage and disadvantage. The proposed weighted D-KNN classifier achieves a recognition rate of more than 80% and outperforms other six classifiers.