Cooperative chameleon search optimization enabled deep ensemble classifier for emotion recognition using EEG signal

The extreme demand for determining the behavioral and cognitive functioning of individuals insists on the need for exploring efficient emotion recognition utilizing the Electroencephalography (EEG) signal. To tackle the issues of extracting the most discriminative features as well as the subject dependency problem of the existing techniques, this research proposes Cooperative chameleon search optimization-enabled deep ensemble Convolutional Recurrent Neural network classifier model (CC-DERN) for effective emotion recognition. The research utilizes the features of extracted statistics, Frequency Domain Features (FDF), Time-Domain Features (TDF), and VGG-16 for enhancing detection accuracy. The technique's importance comes in its ability to combine the tracking, searching, and intelligent hunting elements of the cooperative chameleon search optimization, which offers excellent discriminate feature selection and high detection accuracy for emotion recognition. Additionally, the developed method utilized enhanced preprocessing techniques for removing the artifacts present in the EEG signal that further boosted the detection accuracy. The performance of the CC-DERN technique is analyzed concerning the accuracy, sensitivity, and specificity, which are valued as 96.09%, 99.551%, and 93.430% for the DEAP dataset and 98%, 99.551%, and 84.024% for the SEED dataset, respectively, that show the superiority of the technique with other traditional techniques.