Mining Unlabeled Electronic Music Databases through 3D Interactive Visualization of Latent Component Relationships

Parag Kumar Mital, and Mick Grierson

Proceedings of the International Conference on New Interfaces for Musical Expression

  • Year: 2013
  • Location: Daejeon, Republic of Korea
  • Pages: 227–232
  • Keywords: mir, plca, mfcc, 3d browser, daphne oram, content-based information retrieval, interactive visualization
  • DOI: 10.5281/zenodo.1178614 (Link to paper)
  • PDF link

Abstract:

We present an interactive content-based MIR environment specifically designedto aid in the exploration of databases of experimental electronic music,particularly in cases where little or no metadata exist. In recent years,several rare archives of early experimental electronic music have becomeavailable. The Daphne Oram Collection contains one such archive, consisting ofapproximately 120 hours of 1/4 inch tape recordings and representing a perioddating from circa 1957. This collection is recognized as an importantmusicological resource, representing aspects of the evolution of electronicmusic practices, including early tape editing methods, experimental synthesistechniques and composition. However, it is extremely challenging to derivemeaningful information from this dataset, primarily for three reasons. First,the dataset is very large. Second, there is limited metadata --- some titles,track lists, and occasional handwritten notes exist, but where this is true,the reliability of the annotations are unknown. Finally, and mostsignificantly, as this is a collection of early experimental electronic music,the sonic characteristics of the material are often not consistent withtraditional musical information. In other words, there is no score, no knowninstrumentation, and often no recognizable acoustic source. We present amethod for the construction of a frequency component dictionary derived fromthe collection via Probabilistic Latent Component Analysis (PLCA), anddemonstrate how an interactive 3D visualization of the relationships betweenthe PLCA-derived dictionary and the archive is facilitating researcher'sunderstanding of the data.

Citation:

Parag Kumar Mital, and Mick Grierson. 2013. Mining Unlabeled Electronic Music Databases through 3D Interactive Visualization of Latent Component Relationships. Proceedings of the International Conference on New Interfaces for Musical Expression. DOI: 10.5281/zenodo.1178614

BibTeX Entry:

  @inproceedings{Mital2013,
 abstract = {We present an interactive content-based MIR environment specifically designedto aid in the exploration of databases of experimental electronic music,particularly in cases where little or no metadata exist. In recent years,several rare archives of early experimental electronic music have becomeavailable. The Daphne Oram Collection contains one such archive, consisting ofapproximately 120 hours of 1/4 inch tape recordings and representing a perioddating from circa 1957. This collection is recognized as an importantmusicological resource, representing aspects of the evolution of electronicmusic practices, including early tape editing methods, experimental synthesistechniques and composition. However, it is extremely challenging to derivemeaningful information from this dataset, primarily for three reasons. First,the dataset is very large. Second, there is limited metadata --- some titles,track lists, and occasional handwritten notes exist, but where this is true,the reliability of the annotations are unknown. Finally, and mostsignificantly, as this is a collection of early experimental electronic music,the sonic characteristics of the material are often not consistent withtraditional musical information. In other words, there is no score, no knowninstrumentation, and often no recognizable acoustic source. We present amethod for the construction of a frequency component dictionary derived fromthe collection via Probabilistic Latent Component Analysis (PLCA), anddemonstrate how an interactive 3D visualization of the relationships betweenthe PLCA-derived dictionary and the archive is facilitating researcher'sunderstanding of the data.},
 address = {Daejeon, Republic of Korea},
 author = {Parag Kumar Mital and Mick Grierson},
 booktitle = {Proceedings of the International Conference on New Interfaces for Musical Expression},
 doi = {10.5281/zenodo.1178614},
 issn = {2220-4806},
 keywords = {mir, plca, mfcc, 3d browser, daphne oram, content-based information retrieval, interactive visualization},
 month = {May},
 pages = {227--232},
 publisher = {Graduate School of Culture Technology, KAIST},
 title = {Mining Unlabeled Electronic Music Databases through {3D} Interactive Visualization of Latent Component Relationships},
 url = {http://www.nime.org/proceedings/2013/nime2013_132.pdf},
 year = {2013}
}