Project made by Harshvardhan Vatsa Face recognition using PyTorch,CNN and OpenCV This Project requires the following dependencies 1)PyTorch Machine and Deep Learning Framework 2)Facenet_pytorch- Package that identifies Faces 3)PIL=Python Imaging Library 4)Open CV2 - Open Computer Vision Library 5)Time and OS packages(NOT COMPULSORY BUT Reccomended)
Instructions: ->In the Photos Folder, make a folder of your name and add photos of yourself ->Run The Program and a window called "Face Detection" will pop up, you can view a frame with your name and probablity on it -> press escape button to terminate the window -> press space bar to add the picture in your library for more further accurate results
Additional Notes: This project uses prelearnt dataset called vggface2 which contains around 3.31 million images divided into 9131 classes,each of resolution below 32 pixels. PyTorch uses MTCNN(Multi-task Cascaded Convolutional Networks) to learn the faces in your folder and detect with the help of the vggface2 dataset.