Deep Learning Based Human Emotion Exposure Detection Using Vocal and Countenance

Abstract

The subtle expression of emotions on a person's face might give insight into the ideas running through that person's head. Today's society depends on being able to read facial emotions. Faces are a universal language that humans use to communicate a common and basic set of emotions. In a variety of settings and spheres of life, emotion recognition has benefits. For the purposes of safety and health, it is beneficial and essential. Additionally, it is essential for quickly and simply determining human emotions at a particular time without actually asking. Facial expression with non verbal cues are important for interpersonal interaction. The method of determining a person's emotional state is done using a facial expression recognition system. The technique of identifying different facets of a person's facial expressions is called emotion recognition. One of the most potent and challenging jobs in social communication is identifying human emotion from video frames, or facial expressions. This system compares the captured image to the training dataset that is stored in the database, and then displays the emotional state of the image. Recognizing automatically, the human emotions in pictures and videos will be possible with the help of an algorithm that carries out the identification, extracting and evaluating these facial expressions. These systems allow detection for emotions such as joy, rage, sadness, disgust, surprise, fear, neutral, and more. The applications for DL-based human emotion detection attempt to comprehend the significance of these features of body language and apply this understanding to various datasets and sources of information for the face emotion CNN pretrained module mobilenetv2 employed. This can assist in identifying and extracting facial expressions. These algorithms outperform numerous datasets of photos and videos in terms of accuracy.