Large margin deep embedding for aesthetic image classification  

作　　者：Guanjun GUO Hanzi WANG Yan YAN Liming ZHANG Bo LI 

机构地区：[1]Fujian Key Laboratory of Sensing and Computing for Smart City,School of Information Science and Engineering,Xiamen University,Xiamen 361005,China [2]Faculty of Science and Technology,University of Macao,Macao 999078,China [3]Beijing Key Laboratory of Digital Media,School of Computer Science and Engineering,Beihang University,Beijing 100191,China

出　　处：《Science China(Information Sciences)》2020年第1期225-227,共3页中国科学（信息科学）（英文版）

基　　金：supported by National Natural Science Foundation of China (Grant Nos. U1605252, 61872307, 61472334, 61571379);National Key R&D Program of China (Grant No. 2017YFB1302400);UM Multi-Year Research (Grant No. MYRG201700218-FST)

摘　　要：Dear editor,The problem of aesthetic image classification has attracted much attention during the past few years.The recently proposed methods[1–4]based on the deep convolutional neural network(CNN)[5,6]have achieved large improvements over the methods based on the handcrafted aesthetic features.Although the existing CNN-based methods have shown superior performance,the task of aesthetic image classification is still very challenging.This is mainly due to the fact that aesthetic images are usually captured in complex environments and they have different subjects and styles.

关 键 词：EDITOR IMAGE NEURAL 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] TP183[自动化与计算机技术—计算机科学与技术]