JAMS: A JSON annotated music specification for reproducible MIR research

Eric J. Humphrey, Justin Salamon, Oriol Nieto, Jon Forsyth, Rachel M. Bittner, Juan P. Bello

Research output: Contribution to conferencePaperpeer-review

Abstract

The continued growth of MIR is motivating more complex annotation data, consisting of richer information, multiple annotations for a given task, and multiple tasks for a given music signal. In this work, we propose JAMS, a JSON-based music annotation format capable of addressing the evolving research requirements of the community, based on the three core principles of simplicity, structure and sustainability. It is designed to support existing data while encouraging the transition to more consistent, comprehensive, well-documented annotations that are poised to be at the crux of future MIR research. Finally, we provide a formal schema, software tools, and popular datasets in the proposed format to lower barriers to entry, and discuss how now is a crucial time to make a concerted effort toward sustainable annotation standards.

Original languageEnglish (US)
Pages591-596
Number of pages6
StatePublished - 2014
Event15th International Society for Music Information Retrieval Conference, ISMIR 2014 - Taipei, Taiwan, Province of China
Duration: Oct 27 2014Oct 31 2014

Conference

Conference15th International Society for Music Information Retrieval Conference, ISMIR 2014
Country/TerritoryTaiwan, Province of China
CityTaipei
Period10/27/1410/31/14

ASJC Scopus subject areas

  • Music
  • Information Systems

Fingerprint

Dive into the research topics of 'JAMS: A JSON annotated music specification for reproducible MIR research'. Together they form a unique fingerprint.

Cite this