Transcribed Digits and Apparel Classification using Convolutional Neural Network

Image classification is one of the most trending technologies that have lot many applications. It would not be wrong if we say that it has become a necessity for research community due to the explosion of image data, we deal in our daily life. This paper illustrates and exemplifies the influence of Convolutional Neural Networks (CNN) in the challenging sophisticated image classification riddle. CNN are composed of manifold processing layers with numerous abstract levels. Millions of parameters to tune the model, multiple layers and handcrafted hyperparameters make them unique and significantly important in image processing. In this research paper, our focus has been to explore the structure of CNN along with its various parameters and formulating the problem on the benchmark datasets - MNIST & Fashion-MNIST (FMNIST).