Continuous wavelet transform based speech emotion recognition

Emotion recognition from speech is one of the most interesting topics in research community and has developed to a great extent in the last few years. The real challenge in speech emotion recognition (ER) lies in the extraction of features that efficiently encapsulate the emotional information in speech and also do not depend on the speaker. This paper deals with the challenging task of speaker independent ER based on feature selection and classification algorithms. Features are selected based on continuous wavelet transform (CWT) and prosodic coefficients, and are classified and compared using support vector machine (SVM).