Abstract

Password authentication is readily cracked, and hackers may simply predict the patterns. Faceprint plays a significant role , making it more difficult for hackers to match it up. A customizable app that takes the faceprint as an input, compares it to the current database, and analyses facial attributes to obtain a more exact face pattern to eliminate errors is created. The strategy comprises some modular and efficient algorithms like Siamese Neural Network (twin neural network) for accurate one-shot image recognition. Dlib is utilized for facial landmark detection. Then, using Retina Face for both one face and multiface detection in a crowd. And a pixel wise face localization method, Retina face to perform tasks such as face detection,2D face alignment and 3D face reconstruction simultaneously for accurate precision.The technology leverages a modular design to reduce the need for separate facial recognition systems. It is easily adaptable to all locations by just changing a few lines of code to meet the needs. With the advent of facial recognition technology, the technique will assist in quickly adapting to the changes required by diverse sectors