深度卷积神经网络模型发展综述

Review of the Development of Deep Convolutional Neural Network Model

随着移动互联网与硬件处理器技术的不断发展,海量数据处理与计算能力不断提高,深度学习备受关注。卷积神经网络是深度学习模型中最重要的一种结构,可用于目标特征提取。介绍了为提高卷积神经网络性能,不断增加卷积网络深度的模型,以及因此带来的新问题和解决方法。

With the continuous development of mobile Internet,hardware processor and other aspects,and the continuous improvement of massive data and computing power,deep learning has attracted more and more attention of the world.Especially after Lee Sedol fought against Alphago,it attracted worldwide attention.Convolutional neural network is the most important structure in deep learning model,which is used to extract target features.With the continuous development of the deep learning field,this paper introduces the improvement of the performance of the convolutional neural network,the convolutional network models with increasing depth,as well as the new problems and their solutions.