Real-time speech emotion recognition by minimum number of features

Speech emotion recognition is an aspect of equipped robots with the human capabilities. The need to tradeoff between computational volume and performance accuracy is the main challenge of real-time processes. Application domain of this paper is robotic; therefore both mentioned factors are important. Selecting distinguishing factor with low dimension and high resolution is the optimal solution for both decreasing computational volume and increasing performance accuracy. In this paper a feature vector by minimum number of elements is proposed for recognizing emotional states of speech. The elements of proposed feature vector are selected from both prosodic and frequency features. The result of evaluation realizes the desired performance in the real-time processes.