Results of Using Neural Networks to Automatically Creation Musical Compositions Based on Color Image

  • Vladimir RozalievEmail author
  • Nikita Nikitin
  • Yulia Orlova
  • Alla Zaboleeva-Zotova
Conference paper
Part of the Advances in Intelligent Systems and Computing book series (AISC, volume 875)


In this work we show the results of development and experiments with the program for automated sound generation based on image color spectrum with using the neural network. The work contains a description of the transition between color and music characteristics, the rationale for choosing and the description of a recurrent neural network. The choices of the neural network implementation technology as well as the results of the experiment are described.


Sound generation Emotion Image analysis Recurrent neural network Sampling 



The work is partially supported by the Russian Foundation for Basic Research (16-47-340320 and 17-07-01601 projects).


  1. 1.
    Rozaliev, V., Zaboleeva-Zotova, A.: Methods and Models for Identifying Human Emotions by Recognition Gestures and Motion. Atlantis Press, Singapore (2013)CrossRefGoogle Scholar
  2. 2.
    Ariza, C.: Two pioneering projects from the early history of computer-aided algorithmic composition. Comput. Music. J. 35(3), 40–56 (2012)CrossRefGoogle Scholar
  3. 3.
    Chereshniuk, I.: Algorithmic composition and its role in modern musical education. Art Educ. (3), 65–68 (2015)Google Scholar
  4. 4.
    Vygotsky, L.: Imagination and creativity in childhood. J. Russ. East Eur. Psychol. 42(1), 7–97 (2004)CrossRefGoogle Scholar
  5. 5.
    Cope, D.: Computer Models of Musical Creativity. MIT Press, Cambridge (2005)Google Scholar
  6. 6.
    Mazurowski, L.: Computer models for algorithmic music composition. In: Proceedings of the Federated Conference on Computer Science and Information Systems, pp. 733–737 (2012)Google Scholar
  7. 7.
    Sak, H., Senior, A., Beaufays, F.: Long short-term memory based recurrent neural network architectures for large vocabulary speech recognition. ArXiv e-prints (2014)Google Scholar
  8. 8.
    Doornbusch, P., Nierhaus, G.: Algorithmic composition paradigms of automated music generation. Comput. Music J. 34(3), 70–74 (2014)CrossRefGoogle Scholar
  9. 9.
    Brinkkemper, F.: Analyzing six deep learning tools for music generation. Last Accessed 05 May 2018
  10. 10.
    Mikolov, T.: Recurrent neural network based language model. In: Proceedings of Interspeech International Speech Communication Association, vol. 2010, no. 9, pp. 1045–1048 (2010)Google Scholar
  11. 11.
    Kim, H.K., Ao, S.I., Mahyar, A.: Transactions on Engineering Technologies. Special Issue of the World Congress on Engineering and Computer Science, p. 796. Springer Publishing Company, New York (2013)Google Scholar
  12. 12.
    Fernandez, J.D., Vico, F.: AI methods in algorithmic composition. a comprehensive survey. J. Artif. Intell. Res. 48, 513–582 (2013)MathSciNetCrossRefGoogle Scholar
  13. 13.
    Korvel, G., Simonyte, V., Slivinskas, V.: A modified additive synthesis method using source-filter model. J. Audio Eng. Soc. 63(6), 443–450 (2015)CrossRefGoogle Scholar
  14. 14.
    Lazzarini, V., Timoney, J.: Theory and practice of modified frequency modulation synthesis. J. Audio Eng. Soc. 58, 459–471 (2012)Google Scholar
  15. 15.
    Russ, M.: Sound Synthesis and Sampling, p. 568. Taylor & Francis Group, London (2012)CrossRefGoogle Scholar
  16. 16.
    Nikitin, N.A., Rozaliev, V.L. Orlova, Y.A.: Program for sound generation based on image color spectrum with using the recurrent neural network. In: Proceedings of the IV International research conference «Information technologies in Science, Management, Social sphere and Medicine» (ITSMSSM 2017), vol. 72, pp. 227–232. Atlantis Press (2017)Google Scholar

Copyright information

© Springer Nature Switzerland AG 2019

Authors and Affiliations

  • Vladimir Rozaliev
    • 1
    Email author
  • Nikita Nikitin
    • 1
  • Yulia Orlova
    • 1
  • Alla Zaboleeva-Zotova
    • 1
  1. 1.Volgograd State Technical UniversityVolgogradRussia

Personalised recommendations