Create a Face Recog. Model using VGG16 pre-trained model
Pre-requisites: cv2 module, keras, numpy and pillow library.
Step 1: Created Dataset of images using cv2 module from below code
Step 2: Trained my model by using VGG16 pre-trained model in which i have freezed all layers except last FC layers and created new FC layers using addTopModel function and provided my Dataset
I have reached high accuracy because i have given limited datasets and images were very identical
Step 3: By this our model is trained for our images. Now i used mine image for prediction and it gave perfect result.
So this was my model. For datasets i created both testing and validation datasets and took 400 images for training and 100 for validation for each class using cv2 module.