Abstract

The application of face recognition technology has risen significantly in a variety of areas, including the human- computer interfaces, security and surveillance. However, age variance—the condition whereby an individual’s appearance changes drastically over time, frequently provides challenges with standard face recognition algorithms. In order to achieve accurate and reliable face recognition across multiple ages, this study offers an innovative approach to solve age-resilient face recognition merging the FaceNet deep learning model with the Multi-Tasking Cascaded Convolutional Neural Networks (MTCNN). In order to accomplish successful face matching, the proposed approach takes advantage of the FaceNet model, which collects discriminative features from facial images and maps them to a highdimensional feature space. To help to mitigate the effect of age-related changes in facial geometry, the MTCNN is added as a pre-processing step to accurately identify and align faces, hence resolving the challenge of age-related variations. This method offers a flexible and useful solution for age-resilient face recognition by successfully doing away with the requirement for age-specific databases and age group classification. The system’s precision and durability across a large age range have been shown through extensive trials on benchmark datasets, yielding state-of-the-art results in age-resilient face recognition. By guaranteeing dependable and precise face recognition regardless of an individual’s age, the recommended approach has major potential for boosting security and increasing user experience in applications like access control, personal identification, and support for consumers.