Emotional Evaluation of Facial Tension Using LSTM for Psychological Assessments

Recent years have seen a significant invisible stress during and after the pandemic, making emotion a hot topic. The current economy's inflation and rising cost of living are making this worse; many workers are negatively impacted, which has led to many families experiencing heartbreaking situations and a sharp decline in productivity. Growing stress is detrimental to a person's well-being as well as the development of the business and nation. While using a single model to identify emotion is less accurate, using many models can result in processing delays and potentially misleading findings if the input model data is not correctly filtered and segregated. This paper aims to examine, analyse, and theoretically compare research works on emotion recognition using 15 facial expression methods and 17 voice approaches. It will go over the advantages and disadvantages of each technique as well as the accuracy of certain independent and hybrid techniques for recognizing emotions. While some techniques (such CNN, KNN, and SVM) can be used to many models, their levels of strength vary. Finding this is crucial in order to improve or replace the weaker level when using the same technique on several models. For simplicity of visual comparison, this article will also show the various popularity levels of the procedures in each model. Ideally, it will enable novice researchers to quickly choose the best technique for interpreting a facial expression or vocal cue to identify an emotion.