I have been to build a deep learning model, called ReNet50, for distinguishing the picture. If you are interested in it, you can find it in my Github: https://github.com/zero0827z/Deep-Learning-and-visualize