SCIENCE CHINA Information Sciences, Volume 63, Issue 1: 119101(2020) https://doi.org/10.1007/s11432-018-9567-8

Large margin deep embedding for aesthetic image classification

More info
  • ReceivedMay 25, 2018
  • AcceptedAug 3, 2018
  • PublishedSep 12, 2019


There is no abstract available for this article.


This work was supported by National Natural Science Foundation of China (Grant Nos. U1605252, 61872307, 61472334, 61571379), National Key RD Program of China (Grant No. 2017YFB1302400), and UM Multi-Year Research (Grant No. MYRG2017-00218-FST).




[1] Tang X, Luo W, Wang X. Content-Based Photo Quality Assessment. IEEE Trans Multimedia, 2013, 15: 1930-1943 CrossRef Google Scholar

[2] Datta R, Joshi D, Li J, et al. Studying aesthetics in photographic images using a computational approach. In: Proceedings of European Conference on Computer Vision, 2006. 288--301. Google Scholar

[3] Guo G, Wang H, Shen C. Automatic Image Cropping for Visual Aesthetic Enhancement Using Deep Neural Networks and Cascaded Regression. IEEE Trans Multimedia, 2018, 20: 2073-2085 CrossRef Google Scholar

[4] Pang Y, Wang S, Yuan Y. Learning regularized LDA by clustering.. IEEE Trans Neural Netw Learning Syst, 2014, 25: 2191-2201 CrossRef PubMed Google Scholar

[5] Krizhevsky A, Sutskever I, Hinton G H. ImageNet classification with deep convolutional neural networks. In: Proceedings of the 25th International Conference on Neural Information Processing Systems, 2012. 1097--1105. Google Scholar

[6] Jiang X, Pang Y, Sun M. Cascaded Subpatch Networks for Effective CNNs.. IEEE Trans Neural Netw Learning Syst, 2018, 29: 2684-2694 CrossRef PubMed Google Scholar

[7] Schroff F, Kalenichenko D, Philbin J. Facenet: a unified embedding for face recognition and clustering. In: Proceedings of IEEE Conference on Computer Vision and Pattern Recognition, 2015. 815--823. Google Scholar

[8] Rojas R. Neural Networks: A Systematic Introduction. Berlin: Springer, 1996. Google Scholar

[9] Duchi J, Hazan E, Singer Y. Adaptive subgradient methods for online learning and stochastic optimization. J Mach Learn Res, 2011, 12: 2121--2159. Google Scholar

  • Table 1   The classification accuracy obtained by CAP, CB and the proposed LMDE methods for aesthetic image classification on the CHUKPQ dataset$^{\rm~a)}$
    Method Animal Plant Static Architecture Landscape Human Night Overall
    CAP [2] (%) 78.61 76.38 71.74 73.86 77.53 76.94 64.21 77.92
    CB [1] (%) 89.37 91.82 90.69 92.75 94.68 97.40 84.63 92.09
    LMDE (%) 95.54 96.70 94.93 92.43 96.49 95.35 92.18 94.80

    a) The bold fonts represent the highest classification accuracy.

Copyright 2020 Science China Press Co., Ltd. 《中国科学》杂志社有限责任公司 版权所有