Fuzzy Facial Expression Recognition and Recommendation System

Abstract

We can use Facial Expression Recognition (FER) to detect human behavior. We have several ways of measuring human behavior in various situations, such as hand gestures and others, but facial recognition is the best at the moment because it requires the least amount of hardware intervention. The proposed system detects seven essential facial expressions including fear, surprise, happiness, sadness, disgust, neutrality, and anger. Several more categorizations had also crafted the use of Fuzzy Systems. Fuzzy is now a viable idea for fuzzy classification that could assess several random data to match the information in groups based on partial truth. Its use is the technique to identify the face's different parts and movements. Numerous classifications & control problems, notably FER, have been solved via fuzzy systems. We are using publicly available datasets, as well as the established data selection and valuation methods for these datasets. We describe the FER rules/steps, as well as the accompanying information and ideas for applicable applications at each stage. We provide contemporary image processing and accompanying training methodologies for FER based on both static and dynamic image sequences, as well as their pros and cons, for just the recent in deep FER. The system also can predict the percentage of human behavior. The accuracy of the system is very reliable. We're designing a suggestion system that detects the user's facial expressions and predicts their behavior and suggests things that are both readable and listenable form. The FER evaluates human behavior and compares it with its trained model. And, at the same time, make some recommendations to the user.