INTERNATIONAL JOURNAL OF ADVANCE RESEARCH
IN MULTIDISCIPLINARY

Understanding and detecting human emotions by combining data from several sources, such as speech, text, facial expressions, and physiological signs, is known as multimodal emotion identification. Ubiquitous Computing describes how computers and their apps have become an integral part of modern living. Users' interactions with computers are increasingly resembling those between humans. This study presents a novel deep-learning method based on the BROA to facilitate face emotion identification using several data modalities. The BROA optimization method integrates the Bat method (BA) with the Rider Optimization Algorithm (ROA). This integration seeks to improve the precision of emotion identification by adeptly using facial pictures, EEG data, and physiological signals as components of the multimodal input.