Human Speech and Facial Emotion Recognition Technique Using SVM

Human Speech and Facial are the most significant information carriers for human cognitive-communication and recognizing human's identity and emotional status. With the further growth of computer processing capability and the increase of demand for intelligent living, recognition of emotion based on face and speech became the most significant in the applications of Human-Computer Interaction (HCI). In this paper, Human Speech and Facial based emotion recognition technique using a support vector machine (SVM) has been proposed for improving the performance of detection with multi-emotions effectively. The obtained results of the proposed technique show that the average rate of recognition is higher than other recently existing techniques, and the obtained accuracy is 92.88% for facial model and 85.72 % for speech model with low time-consuming.