Speech Emotion Recognition Using Machine Learning

Speech emotion recognition (SER) is a challenging issue because it is not clear which features are effective for classification. Emotionally related features are always extracted from speech signals for emotional classification. Different persons have different emotions and altogether a different way to express it. 

Speech emotion do have different energies, pitch variations are emphasized if considering different subjects. 

Therefore, the speech emotion recognition is a demanding task in computing vision. Handcrafted features are mainly used for emotional identification from audio signals. However, these features are not sufficient to correctly identify the emotional state of the speaker. 

Here, the speech emotion recognition and the classifiers are used to differentiate emotions such as surprise, anger, sadness, neutral state, happiness, etc. 

The advantages of a convolutional neural network (CNN) are investigated in the proposed work. A neural network framework is used to extract the features from speech emotion databases. In this work, we do the preprocessing for the input audio file and CNN layers adopt for feature selection (FS) approach to find the discriminative and most important features for SER. The proposed method shows the better results in terms of the emotion recognition.

Keywords: - Emotion recognition, Convolutional neural network, Features, Speech, Deep learning.