Facial Recognition based Prediction of Employee Stress using Fuzzy Classification

Abstract

Stress has become increasingly common among employees due to the high demands and long working hours of their jobs. Stress can last for a short or long period of time, but it has a mental impact and can lead to a variety of health problems. Early prediction of stress levels can help organizations take proactive measures to prevent burnout and improve employee well-being. This paper presents Facial Recognition based prediction of Employee Stress using Fuzzy Classification. Cameras mounted on the office monitor which provide the image input for the described method. Fuzzy classifiers are based on fuzzy logic which accounts for the various ranges of possibilities between "yes" and "no" in order to simulate human decision-making. The Gabor filter initially identifies the eyes and lips, which are the most visible human facial features. Subsequently, the program will extract their movement information. Next the stress levels of the employees are calculated. Our system primary focuses on stress management, creating a healthy and spontaneous work environment for employees and maximizing their productivity during working hours. Classification accuracy, precision, False Positive Rate (FPR) and True Positive Rate (TPR) have all been used to evaluate the performance. With an accuracy of 98%, TPR of 97%, FPR of 3% and precision of 99%, the described mental health prediction model for employees is more effective than other models.