Masked face recognition is a biometric authentication method widely utilized in smart devices, such as surveillance and attendance systems. The post-pandemic period has reinforced the use of masks in public places as an effective measure for preventing the spread of infections. Earlier face recognition systems were primarily trained on datasets of unmasked human faces. However, the COVID-19 pandemic has necessitated the adaptation of these programs to effectively identify individuals wearing masks. Detecting masked faces presents significant challenges due to the reduced visibility of key facial features. With much of the face obscured, the process becomes more complex, making it harder for algorithms to accurately identify individuals. In this work, we propose an approach that integrates the distinctive features of deep learning models into a unified framework for masked face recognition. This integration aims to enhance the accuracy and effectiveness of recognizing individuals wearing masks by leveraging the strengths of various deep learning techniques. An ensemble model is proposed where deep learning models (SSD and MTCNN) are used for face detection, and facial features are combined with FaceNet embeddings derived from available landmarks such as the eyes, brows, forehead, and other visible areas. A reproducible set of experiments was conducted on benchmark datasets, achieving an accuracy of approximately 94%, representing a significant improvement over the existing state-of-the-art methods.