Face image classification using a simple neural network with steps of training neural networks.
Data was prepared using the FDDB dataset.The Convolutional Neural Network was made using MNIST tutorial available on tensorflow. It was a simple 3 layer perceptron.I t consisted of 2 convolutional layers, 2 pooling layers and one dense layer. The convolution layers were 32 and 64 5*5 filters and the dense layer contains 2 neurons for the 2 classes(face and non face). The label was set to 0 for a non face image and 1 for a face image.