Analysis of Existing Neural Networks for Image Classification

Abstract

Aspects of image classification are considered. The description of the principles of construction of convolutional neural networks is given. The principle of operation of the basic convolutional neural network is considered. The basic concepts necessary for working with neural networks are studied. A general overview of neural networks for image classification is proposed. Considered such neural networks as: LeNet, AlexNet, VGG image Net, Enet. Particular attention is paid to the architecture of neural networks. Their distinctive properties, advantages and disadvantages of neural networks were highlighted and an analysis of their effectiveness was carried out