A Survey on Affective Computing for Psychological Emotion Recognition

Human emotions are precious and it directly exposes the mental wellbeing of human beings. Understanding the real emotion of people creates clues for many decisions making in critical situations. With recent enhancements in the field of artificial intelligence and machine learning, affective computing becomes an interesting area of researches that adopts the emotional behavior of humans and improves the learning impacts that are closely related to behavioral phycology. Evaluation of machine learning algorithms improves the prediction quality, but the problem arises at a decision making that is closely related to each other. The problem of single and multiple modality issues arises during the final emotion identification that makes the decision resilient. The proposed study focused on a broad survey about different techniques used to detect human behavior and real emotion by considering the data analysis tools and platforms. The proposed study takes affective computing as an initiate, considers the physiological data and their relation in the detection of emotions. The study also considers the recent techniques in comparison like deep learning algorithms, machine learning, and toolsets related to that. To find the compact algorithm that works well in decision making, also helpful in creating a novel idea for emotion recognition even better the proposed study is formulated.