Archive for the ·


· Category...


no comments

Looked for papers in the field. After finding relevant papers, updated the bibliography with these paper (see References)


  • [2002,patent] bibtex
    Logan, Beth and Salomon, Ariel, Music Similarity Function Based on Signal AnalysisHouston, TX (US): , 2002.
    @PATENT{2001LoganSalomon, nationality = {United States},
      number = {US 202/0181711 A1},
      year = {2002},
      yearfiled = {2001},
      author = {Logan, Beth and Salomon, Ariel},
      title = {Music Similarity Function Based on Signal Analysis},
      language = {English},
      assignee = {Compaq Information Technologies Group, L.P.},
      address = {Houston, TX (US)},
      day = {December 5},
      dayfiled = {October 31},
      abstract = {The present invention computer method and apparatus determines music similarity by generating K-means (instead of Gaussian) cluster signature and a beat signature for each piece of music. The beat of the music is included in subsequence distance measurement.},
      file = {:C\:\\Users\\Robert\\Documents\\My Dropbox\\Uni\\Thesis\\Shared\\Papers\\2001LoganSalomon.pdf:PDF},
      owner = {Robert},
      timestamp = {2010.04.06}
  • [2002,conference] bibtex
    M. Cooper and J. Foote, "Automatic Music Summarization via Similarity Analysis," in ISMIR 2002, 2002.
      author = {Matthew Cooper and Jonathan Foote},
      title = {Automatic Music Summarization via Similarity Analysis},
      booktitle = {ISMIR 2002},
      year = {2002},
      volume = {October 13, 2002},
      publisher = {IRCAM – Centre Pompidou},
      abstract = {We present methods for automatically producing summary excerpts or thumbnails of music. To find the most representative excerpt, we maximize the average segment similarity to the entire work. After window-based audio parameterization, a quantitative similarity measure is calculated between every pair of windows, and the results are embedded in a 2-D similarity matrix. Summing the similarity matrix over the support of a segment results in a measure of how similar that segment is to the whole. This measure is maximized to find the segment that best represents the entire work. We discuss variations on the method, and present experimental results for orchestral music, popular songs, and jazz. These results demonstrate that the method finds significantly representative excerpts, using very few assumptions about the source audio.},
      file = {:C\:\\Users\\Robert\\Documents\\My Dropbox\\Uni\\Thesis\\Shared\\Papers\\2002CooperFoote.pdf:PDF},
      journal = {2002 International Symposium on Music Information Retrieval},
      owner = {Robert},
      timestamp = {2010.04.06}
  • [2003,article] bibtex
    R. Typke, F. Wiering, and R. C. Veltkamp, "Transportation distances and human perception of melodic similarity," Musicae Scientiae, vol. 4a, pp. 153-181, 2003.
      author = {Rainer Typke and Frans Wiering and Remco C. Veltkamp},
      title = {Transportation distances and human perception of melodic similarity},
      journal = {Musicae Scientiae},
      year = {2003},
      volume = {4a},
      pages = {153--181},
      abstract = {Most of the existing methods for measuring melodic similarity use one-dimensional textual representations of music notation, so that melodic similarity can be measured by calculating editing distances. We view notes as weighted points in a two-dimensional space, with the coordinates of the points reflecting the pitch and onset time of notes and the weights of points depending on the corresponding notes’ duration and importance. This enables us to measure similarity by using the Earth Mover’s Distance (EMD) and the Proportional Transportation Distance (PTD), a pseudo-metric for weighted point sets which is based on the EMD. A comparison of our experiment results with earlier work shows that by using weighted point sets and the EMD/PTD instead of Howard’s method (1998) using the DARMS encoding for determining melodic similarity, it is possible to group together about twice as many known occurrences of a melody within the RISM A/II collection. Also, the percentage of successfully identified authors of anonymous incipits can almost be doubled by comparing weighted point sets instead of looking for identical representations in Plaine & Easie encoding as Schlichte did in 1990.},
      file = {:C\:\\Users\\Robert\\Documents\\My Dropbox\\Uni\\Thesis\\Shared\\Papers\\2003Typke.pdf:PDF}
  • [2004,article] bibtex Go to document
    A. Berenzweig, B. Logan, D. P. W. Ellis, and B. Whitman, "A Large-Scale Evaluation of Acoustic and Subjective Music-Similarity Measures," Computer Music Journal, vol. 28, iss. 2, pp. 63-76, 2004.
      author = {Berenzweig, Adam and Logan, Beth and Ellis, Daniel P. W. and Whitman, Brian},
      title = {A Large-Scale Evaluation of Acoustic and Subjective Music-Similarity Measures},
      journal = {Computer Music Journal},
      year = {2004},
      volume = {28},
      pages = {63--76},
      number = {2},
      copyright = {Copyright © 2004 The MIT Press},
      file = {:C\:\\Users\\Robert\\Documents\\My Dropbox\\Uni\\Thesis\\Shared\\Papers\\2004Berenzweig.pdf:PDF},
      issn = {01489267},
      jstor_articletype = {primary_article},
      jstor_formatteddate = {Summer, 2004},
      jstor_issuetitle = {Music Information Retrieval},
      publisher = {The MIT Press},
      url = {}
  • [2006,article] bibtex
    G. Aloupis, T. Fevens, S. Langerman, T. Matsui, A. Mesa, Y. Nuñez, D. Rappaport, and G. Toussaint, "Algorithms for Computing Geometric Measures of Melodic Similarity," Comput. Music J., vol. 30, iss. 3, pp. 67-76, 2006.
      author = {Aloupis, Greg and Fevens, Thomas and Langerman, Stefan and Matsui, Tomomi and Mesa, Antonio and Nu\&\#x00f1;ez, Yurai and Rappaport, David and Toussaint, Godfried},
      title = {Algorithms for Computing Geometric Measures of Melodic Similarity},
      journal = {Comput. Music J.},
      year = {2006},
      volume = {30},
      pages = {67--76},
      number = {3},
      address = {Cambridge, MA, USA},
      citeseerurl = {},
      doi = {},
      file = {:C\:\\Users\\Robert\\Documents\\My Dropbox\\Uni\\Thesis\\Shared\\Papers\\2006Aloupis.pdf:PDF},
      issn = {0148-9267},
      publisher = {MIT Press}
  • [2007,mastersthesis] bibtex
    N. Mitchell, "Music Similarity Metrics: Recognizing Tempo, Transposition, Ornamentation, and Accentuation Properties," Master’s Dissertation , 2007.
      author = {Nicole Mitchell},
      title = {Music Similarity Metrics: Recognizing Tempo, Transposition, Ornamentation, and Accentuation Properties},
      school = {School of Computing, Queen's University. Kingston, Ontario, Canada.},
      year = {2007},
      month = {January},
      file = {:C\:\\Users\\Robert\\Documents\\My Dropbox\\Uni\\Thesis\\Shared\\Papers\\2007Mitchell.pdf:PDF},
      owner = {Robert},
      timestamp = {2010.04.06}


no comments

“How can we build a computational model to determine the probability that one melodic sequence has been derived from another existing melody?”

I’ve intentionally said “determine the probability” because in terms of copyright/intellectual property it can always be debatable as to whether a melody has been ‘borrowed and modified’ or it has been created solely by the composer.
As I see it, it’s not possible to have an algorithm/model make a “complete” decision as to determining that a melody has been borrowed and modified.

With the “Kookaburra vs Down Under” case, it’s fairly obvious that the flute has been derived from the kookaburra song, but it as the article says: “Justice Jacobson said perhaps the clearest illustration of the objective similarity between the songs was Hay’s ‘frank admission of a causal connection between the two melodies and the fact that he sang the relevant bars of Kookaburra when performing Down Under at gigs from 2002.”