Image Classification in the Era of Deep Learning

The use of deep learning has taken immense step in today’s technology, with profound use of Artificial Neural Network. The notable part of Deep learning is the formulation and popularity of a specific Neural Network (NN) known as the Convolutional Neural Networks (CNN or ConvNet). CNN is classified as a feed-forward ANN, used mainly to examine visual images. The problem discussed here is how an image classifier can be implemented using CNN. This proposed work is carried out to give a complete insight on CNN and not regarding it as a black box. This paper mainly concentrates on CNN concepts and implements a simple Multi Image Classification using CNN with the help of Tensor flow, Keras and Theano deep learning libraries. A CNN model was built from scratch and was trained with a dataset. Different Loss functions for Multi Image classification were also implemented.