Automatic Facial Expression Recognition and Classification Using Deep Learning Models

Abstract

Computer security, neuroscience, psychology, and engineering are just few of the fields that can benefit from facial expression recognition, often known as FER. Because it does not intrude on a person's privacy, many people believe it could be an effective tool in the fight against criminal activity. Despite this, FER suffers from a number of drawbacks, the most significant of which is the lack of accuracy of its predictions when applied to severe head postures. An intriguing topic of research, automatic emotion recognition based on facial expression has been presented and implemented in a variety of fields, including safety, health, and in human machine interactions, to name a few of these domains. Researchers in this topic are interested in developing methods to understand facial expressions, create codes for them, and extract these elements so that computers can make more accurate predictions. The extraordinary success of deep learning has led to the method's various sorts of architectures being utilized in an effort to improve performance in order to compete with other approaches. An investigation into new research on automatic facial emotion recognition (FER) using deep learning is going to be the focus of this particular piece of writing. We highlight the contributions that were treated, the architecture, and the databases that were used, and we illustrate the progress that has been made by comparing the approaches that were proposed and the results that were produced. The purpose of this publication is to assist and direct future researchers by offering an overview of recent studies as well as insights that can be used to achieve advancements in this subject. Researchers have become more interested in the field of facial expression recognition over the past decade as it has a wide variety of possible applications. Expression analysis includes a number of crucial steps, one of which is feature extraction, which helps toward the goal of accurate and speedy expression recognition. Expressions of joy, surprise, contempt, sadness, rage, and terror can be shown on their faces. Facial expressions can reveal a person's emotions. The most popular method for determining a person's state of mind is by analyzing their facial expressions. There is a wide range of distinct feelings that can be divided into two groups: happy emotions and negative emotions. There are four primary categories of commonly used systems: face detection and extraction, face classification, face recognition, and face recognition. In the current system, it is not so easy to pinpoint the precise emotion that a person is feeling. in addition to the categorization of frame-based expression recognition We should work on detecting facial expressions and emotions in both good and negative pictures, and We should also create strong systems. As a result, increasing the recognition accuracy is intended to be the end result of this study.