SCIENCE CHINA Information Sciences, Volume 62 , Issue 2 : 027101(2019) https://doi.org/10.1007/s11432-018-9692-2

Two open-source projects for image aesthetic quality assessment

More info
  • ReceivedSep 8, 2018
  • AcceptedNov 14, 2018
  • PublishedDec 26, 2018


There is no abstract available for this article.


[1] Jin X, Chi J Y, Peng S W, et al. Deep image aesthetics classification using inception modules and fine-tuning connected layer. In: Proceedings of the 8th International Conference on Wireless Communications and Signal Processing (WCSP), Yangzhou, 2016. Google Scholar

[2] Jin X, Wu L, Li X. ILGNet: inception modules with connected local and global features for efficient image aesthetic quality classification using domain adaptation. CrossRef Google Scholar

[3] Jin X, Wu L, Li X D, et al. Predicting aesthetic score distribution through cumulative Jensen-Shannon divergence. In: Proceedings of AAAI Conference on Artificial Intelligence, New Orleans, 2018. Google Scholar

[4] Wang J, Lu Y, Liu J. A robust three-stage approach to large-scale urban scene recognition. Sci China Inf Sci, 2017, 60: 103101 CrossRef Google Scholar

  • Figure 1

    (Color online) The predicted aesthetic score histogram according to CJS, RS-CJS, and other loss functions. The leftmost column shows each test image, and the number at the top of each graph is the average value of the scores calculated by the histogram. The second column represents the real image fraction distribution. The third and fourth columns are the results based on the CJS method and the RS-CJS method, respectively. The other columns represent the results of other loss functions.