IoT-Based Face Recognition with SNN

This idea presents a face recognition (FER) algorithm using a deep Siamese neural network (SNN) that stores local patterns of images in a similar space. We create networks to show similarity between pairs of ideas by comparing features with design. Additionally, we developed a new image matching (i.e., positive-negative matching) technique to introduce our Siamese model. Our Siamese model has a proven and accepted model for learning in an integrated environment. The validation method reduces the class variance by reducing the distance between features drawn from the same class, whereas a validation procedure raises the class variance by widening the gap between characteristics drawn from different classes. By using four layers of IoT we successfully implemented the multimodal authentication system with the Biometric and Face Recognition System as a mode for the authentication. In this project machine learning with IoT to make fulfill the product requirements and a more secure authentication system.