By Mukul Kumar
Face Recognition System using well know MTCNN algorithm. This system can be used in various applications such as computer vision, security purposes, etc.
This Face Recognition System is developed in Python language and the outputs are verified using the Google Colaboratory tool. The algorithm used for Face Recognition is MTCNN or Multi-Task Cascaded Convolutional Neural Networks is a neural network that detects faces and facial landmarks on images. MTCNN is one of the most popular and most accurate face detection tools today.
I have initially implemented all the necessary libraries required to perform various tasks such as array manipulation, image processing, mathematical operations, etc.
Here is the description of each function along with the task performed by it-
face_det_ext() to detect the face pixels from entire image
pre_proc() to perform pre-processing tasks on image
get_model_scores() to get the model sores of face input
take_photo() to create a display and capture button interface for capturing a real-time photo
register() to register the initial input photos in the database (these photos will be used for identification of registered user)
scan() to get the real-time input from the user to display the required outputs according to the received input image
Face images of users for initial registration (these images will be saved in the database)
Input face image of a person to be during the working phase
If the system recognizes the input face image is matched with one of the registered user's face image then it will show Hi UserName
If the input image does not match with any of the registered user's face images then it will show Wait a minute! Who are you?