A Comparative Analysis of Machine Learning Algorithms Used For Training in Face Recognition
International Journal of Advanced Trends in Computer Science and Engineering · 2020
To human identity, the face is the most important one that best identifies an individual.Face recognition is very common in today's modern world, and is widely used for much surveillance, safety, retail, tourism, healthcare, and hospitality applications.In the world of facial detection there have been several developments over the years.In this manuscript, the proposed algorithms are likened to the K nearest neighbor, Support Vector Machine, Decision Tree Classifier, Random Forest, Naïve Bayes algorithms.Despite several benefits it provides, the precision of facial recognition may be further increased in cases where the conditions for image processing are not so good and where the photographs of a face are taken many years apart.The accuracy of facial recognition varies according to the technique used and the circumstances under which they are evaluated.The objective of this manuscript is to compared with different machine learning algorithms used for classification depending upon accuracy, precision, sensitivity, specificity.The outcome reveals that all the other algorithms except SVM and KNN fell below 90 per cent in the metrics while the reliability of the test set tends to be reasonably successful for all the algorithms.It was observed that picture face orientation plays a major role in correct face recognition.Both measurements are decreased by 40 percent without face synchronization of the images and hit 60 percent for SVM and even less for other algorithms.