Feature Discovery by Deep Learning for Aesthetic Analysis of Evolved Abstract Images

  • Allan CampbellEmail author
  • Vic Ciesielksi
  • A. K. Qin
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 9027)


We investigated the ability of a Deep Belief Network with logistic nodes, trained unsupervised by Contrastive Divergence, to discover features of evolved abstract art images. Two Restricted Boltzmann Machine models were trained independently on low and high aesthetic class images. The receptive fields (filters) of both models were compared by visual inspection. Roughly 10 % of these filters in the high aesthetic model approximated the form of the high aesthetic training images. The remaining 90 % of filters in the high aesthetic model and all filters in the low aesthetic model appeared noise like. The form of discovered filters was not consistent with the Gabor filter like forms discovered for MNIST training data, possibly revealing an interesting property of the evolved abstract training images. We joined the datasets and trained a Restricted Boltzmann Machine finding that roughly 30 % of the filters approximate the form of the high aesthetic input images. We trained a 10 layer Deep Belief Network on the joint dataset and used the output activities at each layer as training data for traditional classifiers (decision tree and random forest). The highest classification accuracy from learned features (84 %) was achieved at the second hidden layer, indicating that the features discovered by our Deep Learning approach have discriminative power. Above the second hidden layer, classification accuracy decreases.


Computational aesthetics Deep learning Evolved abstract images 


  1. 1.
    Birkhoff, G.D.: Aesthetic Measure. Mass, Cambridge (1933)CrossRefzbMATHGoogle Scholar
  2. 2.
    Campbell, A., Ciesielski, V., Trist, K.: A self organizing map based method for understanding features associated with high aesthetic value evolved abstract images. In: 2014 IEEE Congress on Evolutionary Computation (CEC), pp. 2274–2281. IEEE (2014)Google Scholar
  3. 3.
    Ciesielski, V., Barile, P., Trist, K.: Finding image features associated with high Aesthetic value by machine learning. In: Machado, P., McDermott, J., Carballal, A. (eds.) EvoMUSART 2013. LNCS, vol. 7834, pp. 47–58. Springer, Heidelberg (2013) CrossRefGoogle Scholar
  4. 4.
    Datta, R.: Semantics and aesthetics inference for image search: statistical learning approaches. Pennsylvania State University (2009)Google Scholar
  5. 5.
    Datta, R., Joshi, D., Li, J., Wang, J.Z.: Studying Aesthetics in photographic images using a computational approach. In: Leonardis, A., Bischof, H., Pinz, A. (eds.) ECCV 2006. LNCS, vol. 3953, pp. 288–301. Springer, Heidelberg (2006) CrossRefGoogle Scholar
  6. 6.
    Erhan, D., Bengio, Y., Courville, A., Vincent, P.: Visualizing higher-layer features of a deep network. Dept. IRO, Université de Montréal, Technical report (2009)Google Scholar
  7. 7.
    Fischer, A., Igel, C.: Training restricted boltzmann machines: An introduction. Pattern Recogn. 47(1), 25–39 (2014)CrossRefGoogle Scholar
  8. 8.
    Galanter, P.: Computational aesthetic evaluation: past and future. In: McCormack, J., d’Inverno, M. (eds.) Computers and Creativity, pp. 255–293. Springer, Heidelberg (2012)CrossRefGoogle Scholar
  9. 9.
    Ginosar, S., Haas, D., Brown, T., Malik, J.: Detecting people in cubist art. arXiv preprint arXiv:1409.6235 (2014)
  10. 10.
    Hall, M., Frank, E., Holmes, G., Pfahringer, B., Reutemann, P., Witten, I.H.: The weka data mining software: an update. ACM SIGKDD Explor. Newslett. 11(1), 10–18 (2009)CrossRefGoogle Scholar
  11. 11.
    Hinton, G.: A practical guide to training restricted Boltzmann machines. Momentum 9(1), 926 (2010)Google Scholar
  12. 12.
    Hinton, G., Osindero, S., Teh, Y.-W.: A fast learning algorithm for deep belief nets. Neural Comput. 18(7), 1527–1554 (2006)CrossRefzbMATHMathSciNetGoogle Scholar
  13. 13.
    Geoffrey, E.: Training products of experts by minimizing contrastive divergence. Neural Comput. 14(8), 1771–1800 (2002)CrossRefzbMATHGoogle Scholar
  14. 14.
    Hinton, G.E., Srivastava, N., Krizhevsky, A., Sutskever, I., Salakhutdinov, R.R.: Improving neural networks by preventing co-adaptation of feature detectors. arXiv preprint arXiv:1207.0580 (2012)
  15. 15.
    Hoenig, F.: Defining computational aesthetics. In: Neumann, L., Sbert, M., Gooch, B., Purgathofer, W. (eds.) Computational Aesthetics, pp. 13–18. Eurographics Association, London (2005)Google Scholar
  16. 16.
    Ke, Y., Tang, X., Jing, F.: The design of high-level features for photo quality assessment. In: 2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, vol. 1, pp. 419–426. IEEE (2006)Google Scholar
  17. 17.
    LeCun, Y., Cortes, C.: The mnist database of handwritten digits (1998)Google Scholar
  18. 18.
    Lee, H., Ekanadham, C., Ng, A.Y.: Sparse deep belief net model for visual area v2. In: Platt, J.C., Koller, D., Singer, Y., Roweis, S. (eds.) Advances in Neural Information Processing Systems, pp. 873–880. MIT Press, Cambridge (2008)Google Scholar
  19. 19.
    Lu, X., Lin, Z., Jin, H., Yang, J., Wang, J.Z.: Rapid: Rating pictorial aesthetics using deep learning. In: Proceedings of the ACM International Conference on Multimedia, pp. 457–466. ACM (2014)Google Scholar
  20. 20.
    Machado, P., Cardoso, A.: Generation and evaluation of artworks. In: Proceedings of the 1st European Workshop on Cognitive Modeling, CM’96, pp. 96–39 (2010)Google Scholar
  21. 21.
    Murray, N., Marchesotti, L., Perronnin, F.: Ava: A large-scale database for aesthetic visual analysis. In: 2012 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 2408–2415. IEEE (2012)Google Scholar
  22. 22.
    Nair, V., Hinton, G.E.: Rectified linear units improve restricted boltzmann machines. In: Proceedings of the 27th International Conference on Machine Learning (ICML 2010), pp. 807–814 (2010)Google Scholar
  23. 23.
    Reaves, D.: Aesthetic image rating (AIR) algorithm. Ph.D. thesis (2008)Google Scholar
  24. 24.
    Simard, P.Y., Steinkraus, D., Platt, J.C.: Best practices for convolutional neural networks applied to visual document analysis. In: 2013 12th International Conference on Document Analysis and Recognition, vol. 2, pp. 958–958. IEEE Computer Society (2003)Google Scholar
  25. 25.
    Spratt, E.L., Elgammal, A.: Computational beauty: Aesthetic judgment at the intersection of art and science. arXiv preprint arXiv:1410.2488 (2014)
  26. 26.
    Jost Tobias Springenberg and Martin Riedmiller. Improving deep neural networks with probabilistic maxout units. arXiv preprint arXiv:1312.6116 (2013)
  27. 27.
    Xu, Q., D’Souza, D., Ciesielski, V.: Evolving images for entertainment. In: Proceedings of the 4th Australasian Conference on Interactive Entertainment, p. 26. RMIT University (2007)Google Scholar

Copyright information

© Springer International Publishing Switzerland 2015

Authors and Affiliations

  1. 1.School of Computer Science and Information TechnologyRMIT UniversityMelbourneAustralia

Personalised recommendations