An Effective Speech Emotion Recognition Using Artificial Neural Networks

The speech signal has emerged as the quickest and the most normal mode of communication between the human beings. As an indispensable method of the human emotional behaviour comprehension, the speech emotion recognition (SER) has evinced zooming significance in the human-centred signal processing. In the current document, the speech signal is identified and distinguished by way of four distinct phases such as the feature extraction; modified cuckoo search-based generating finest weight, feature analysis and the artificial neural network (ANN) classification. The speech signal categorization, in turn, is performed in the working platform of MATLAB, and outcomes are assessed to ascertain the efficiency in the performance of the system. The performance metrics are analysed and made comparison for an improvement of the sensitivity 0.692%, specificity 24% and accuracy 5.88% with the existing method.