Improving of Feature Selection in Speech Emotion Recognition Based-on Hybrid Evolutionary Algorithms

One of the important issues in speech emotion recognizing is selecting of appropriate feature sets in order to improve the detection rate and classification accuracy. In last studies researchers tried to select the appropriate features for classification by using the selecting and reducing the space of features methods, such as the Fisher and PCA. In this research, a hybrid evolutionary algorithm is proposed which uses support vector machine classifier and cuckoo search algorithm in combination with the genetic algorithm in order to select the optimal features. In the proposed method, at first, a set of characteristics based on the Cepstral, Spectral and Fourier coefficients of the speech signal is extracted and then with the proposed hybrid algorithm, the operation of selecting the optimal feature set is performed. The results of the experiments on the famous Berlin's emotional speech database showed that using this proposed method for selecting the features, increases the classification accuracy to about 93%.