Novel Approaches in Machine Learning for Enhanced Facial Recognition Systems

Face recognition (FR) is the latest method of authentication and authorization, and it has advanced significantly in the previous decade, with such systems becoming widespread in industries such as security, commerce, and so on. Because of this, FR has become one of the most investigated fields, numerous algorithms having been developed over the years by academic and commercial researchers. Since most existing methods perform poorly in unrestricted environments, this area of study continues to expand rapidly. In this research, a new method for boosting FR accuracy called Gray Wolf Optimization-based AdaBoost (GWO-AdaBoost) is provided. Images of people's faces have been chosen from the OLR Database. The acquired images are then sent to the processing phase, where they are transformed from their raw form into the required form. Gabor Filter (GF) is used to retrieve features from the processed images, and recognition is done using the suggested Machine Learning (ML) model, GWO-AdaBoost. The results of the proposed ML model are compared to those obtained using the AdaBoost method. The results demonstrate that the proposed model outperforms AdaBoost in terms of recognition rate. The impact of input sample count on execution time was also examined. The proposed method takes more time than AdaBoost to identify a person's face.