gavisharora / facerecognition Goto Github PK
View Code? Open in Web Editor NEWPython project based on facial recognition using OpenCV and Histogram of oriented gradients (HOG) model, it can also be modified to run using Convulutional neural networks (CNN) to increase accuracy, HOG was preferred for its ability to run faster on low end devices without GPUs. Place the training data inside the 'known' folder, which has folders labeled as names of people it contains the images of, this is the training data, the test data/images are supposed to be in the 'unknown' folder. The output will be the image and if detected the name of the person detected along with it.