Retraction Notice: Deep Ensemble Learning Model for Face Recognition using Smart Phone

Mobility necessitates a significant degree of capture condition variability, which is difficult to manage and has a direct impact on extraction of facial characteristics and face detection. The much more intriguing option for automatic facial recognition seems to be deep learning, although these systems heavily rely on the model created during the training phase. The scale of the models also makes it challenging to integrate them into apps smartphone, especially when the model needs to be incorporated. In this study, a deep learning ensemble model was trained to recognize faces on low-resource devices and tested for precision, size, and timing to produce quantitative results. Since the goal of this study is to include so many more scenarios as possible, a variety of databases, including both public and private datasets designed to replicate the complexity of mobile scenarios, were used. In order to conduct a fair comparison, publicly accessible models and conventional methodologies were also included in the study. Additionally, the assessment is supplemented with various classifiers and detectors due to the importance of the template matching and face identification stages. Finally, a Java-based Android application is created and tested in order to gather performance information on the entire system that is built into a mobile device.