An efficient emotion recognition system from multimodal signals with cognitive based personality trait mapping using machine learning techniques

Abstract

Emotions and Personality Traits are two inseparable cognitive newlineprocesses in human existence, which have both become integral parts of newlinemodern culture as a response to several reasons. Emotions have a critical role newlinein human behaviour, influence systems such as observation, cognition, newlinedecision-making process, education, and intelligence. This has created an newlineoverriding need to provide an efficient means to handle the Emotion newlineRecognition methodologies from various bio-signal such as EEG signal, newlinePeripheral signal, physiological signal and video for effective understanding of newlinehuman behaviour, cognition, and ability. Humans frequently transmit emotions newlineand their present tailored to meet the needs state through extraneous physical newlineexpressions such as a grin or more physiological responses such as increased newlineirregular heartbeat (HR). newlineAutomatic recognition of human emotion is a challenging job due to newlinefluctuations in the human brain signals at different levels of recognition newlineprocess. In this context, recognition of all five primary human emotions (happy, newlinesad, calm, fear, and angry) from the various modalities of recorded incomplete newlinesignal is an important research problem and receives a growing attention. newlineThe methods and techniques developed so far for emotion newlinerecognition system cannot be applied directly because the collected newlinecomplete/incomplete signal from the stimuli responses are more complex. newlineEmotion recognition mapping to personality trait is complex due to the absence newlineof pure EEG signals along with the minimum number of channels required to newlineidentify the emotion. newline