Convolutional Neural Network (CNN) is one of the most important algorithms used for computer vision task. CNN integrates feature extraction process along with classification. It can automatically extract features from images. Feature extraction is done through convolution operation in CNN. CNN also employs pooling operation for dimensionality reduction. Over the years researchers have proposed different architectures of CNN. The architectures differ based on the numbers of convolution layers, size of filters, number of pooling layers and activation functions used in the CNN model. LeNet 5, AlexNet, VGG16, VGG19 are some of the CNN architectures discussed here