Quality Assurance of Facial Recognition Software for Arabic Social Media Celebrities

In this study, we developed a system to test the quality assurance of face detection for Arabic celebrities in social media using the Haar detection technique with the OpenCV Python package. In our developed system, we used 64 images to determine and find the quality measures, like the error rate of recognizing faces in the images, by employing various testing methods. Our work resulted in creating a data set called Yasmin Nadia Arabc Social Media Images containing faces for Arabic social media celebrations, which is available for public use on the websites Kaggle and GitHub. The developed system can identify faces with an accuracy rate of more than 90% in most cases.