• Vaishnavi Vishwasrao
• Anjali Pandit
• Ritesh Nikalje
• Neha Vishwakarma
• Trupti Khose

The realm of facial recognition technology is witnessing a surge in advancements, thanks to the strides made in deep learning methodologies and the availability of expansive training datasets. Nevertheless, its practical implementation for authentication purposes encounters hurdles in real-world settings marked by variations in facial poses, angles, lighting conditions, and potential obstructions. To tackle these obstacles, this paper introduces an innovative deep learning API that amalgamates the functionalities of MTCNN and FaceNet. By leveraging MTCNN for precise face detection and FaceNet for extracting facial features, the API employs a two-step verification process. FaceNet facilitates the mapping of facial images into a compact Euclidean space, where spatial distances directly correlate with facial similarity, subsequently storing this data securely within a MongoDB database. Utilizing the Deep Face model developed by Facebook's AI research team, the system achieves remarkable accuracy in user verification. This API not only enhances user authentication but also optimizes the user experience, thereby mitigating the risk of unauthorized access. While acknowledging areas for potential improvement, this paper offers a comprehensive summary of its findings and delves into the challenges lying ahead.

Facial recognition, deep learning, MTCNN, FaceNet, MongoDB, deep authentication, Deep Face model, user verification.

"Face Authentication using MTCNN and FaceNet", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.11, Issue 4, page no.g160-g165, April-2024, Available :http://www.jetir.org/papers/JETIR2404616.pdf

"Face Authentication using MTCNN and FaceNet", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.11, Issue 4, page no. ppg160-g165, April-2024, Available at : http://www.jetir.org/papers/JETIR2404616.pdf