• DocumentCode
    3106277
  • Title

    A Proposal for Intervention by User in Interactive Genetic Algorithm for Creation of Music Melody

  • Author

    Koga, Shunsuke ; Inoue, Takeru ; Fukumoto, Masaaki

  • Author_Institution
    Grad. Sch. of Eng., Fukuoka Inst. of Technol., Fukuoka, Japan
  • fYear
    2013
  • fDate
    5-7 July 2013
  • Firstpage
    129
  • Lastpage
    132
  • Abstract
    In the area of Interactive Evolutionary Computation (IEC), some previous studies have already proposed a method for creating music melody. However, in these methods, user´s concrete intention or Kansei seems not to be reflected to the created melodies. This study focuses on this point: reflection of user´s concrete intention and Kansei in IEC. While most of conventional IEC approaches accepted only user´s evaluation by scoring or comparing solution candidates, the proposed method accepts user´s operation for solution candidates as user´s intervention. To investigate the proposed method for creation of music melody, listening experiments were conducted. In the experiment, the subjects listened to music melody created from the IEC system based on the proposed method. With the system, the subjects can control key of music note in melody by themselves when they wanted to change the melody. As result of experiment, the mean fitness value gradually increased in accordance with progress of generation, and significant differences between the first generation and other generations were observed. The experiment also investigated the efficacy of the user´s intervention.
  • Keywords
    genetic algorithms; music; IEC system; Kansei; interactive evolutionary computation; interactive genetic algorithm; music melody creation; music note key control; user intervention; Data visualization; Evolutionary computation; Genetic algorithms; IEC; Music; Sociology; Statistics; genetic algorithm; interactive evolutionary computation; intervention; kansei; music melody;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Biometrics and Kansei Engineering (ICBAKE), 2013 International Conference on
  • Conference_Location
    Tokyo
  • Type

    conf

  • DOI
    10.1109/ICBAKE.2013.26
  • Filename
    6603488