Deep Hunting Social Optimization Algorithm Enabled Deep Neuro Fuzzy Network for Facial Expression Recognition

Recently, a number of researchers have been able to identify human emotions from facial expressions. Realistically speaking, emotion communication is a transient and multimodal process, which makes identifying human emotions difficult. This paper proposes an effective model for multimodal facial expression recognition using the Deer Hunting Social Optimization Algorithm (DHSOA) based Deep Neuro Fuzzy Network (DNFN) in order to overcome such problems in the existing models and achieve outstanding performance in the field of human facial expression recognition. In this work, input video and input electroencephalogram (EEG) signals are considered for additional processing. The feature selection stage uses DHSOA to select appropriate features from input video and EEG signal phases. In contrast, the suggested DHSOA-based DNFN for facial video has a maximum accuracy of 0.910, a precision of 0.899, and a recall of 0.900.