基于深度学习的油画分类网络模型  

A network model for oil painting classification based on deep learning

作　　者：曹建收 陈光喜[2] 任夏荔 康春生 CAO Jianshou;CHEN Guangxi;REN Xiali;KANG Chunsheng(1.School of Computer and Information Security,Guilin University of Electronic Technology;Guangxi Colleges and Universities Key Laboratory of Intelligent Processing of Computer Images and Graphics,Guilin University of Electronic Technology; Guilin 541004,Chin)

机构地区：[1]桂林电子科技大学计算机与信息安全学院,广西桂林541004 [2]桂林电子科技大学广西高校图像图形智能处理重点实验室,广西桂林541004

出　　处：《桂林电子科技大学学报》2018年第1期65-68,共4页Journal of Guilin University of Electronic Technology

基　　金：广西学位与研究教育改革项目(JGY2014060);桂林电子科技大学研究生教育创新计划(2016XWYJ109)

摘　　要：针对既有的深度学习网络结构模型在油画数据集上取得的分类结果较差的问题,提出了一种基于深度学习的油画分类网络模型。利用AlexNet网络抑制过拟合的Dropout算法和GoogleNet网络加深网络深度和宽度的Inception模块,构建一种层数和训练用时相对较少的网络模型。实验结果表明,新网络模型的油画分类准确率比AlexNet和GoogleNet高。In view of the bad classification results of the existing deep learning network structure model in the oil painting data sets,a new network structure model is proposed.By taking advantages of Dropout algorithm of AlexNet network to suppress the overfitting and the inception module of GoogleNet network to deepen the network depth and width.A new network structure model with fewer layers and less training time is constructed.The final results show that the newly constructed network model is more accurate than AlexNet and GoogleNet in oil painting classification.

关 键 词：深度学习 分类 油画 网络结构模型 艺术领域 

分 类 号：TP399[自动化与计算机技术—计算机应用技术]