Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

JAZZVAR: A Dataset of Variations found within Solo Piano Performances of Jazz Standards for Music Overpainting

Eleanor Row, Jingjing Tang, George Fazekas

TL;DR

This paper outlines the curation process for obtaining and sorting the repertoire, the pipeline for creating the Original and Variation pairs, and the analysis of the JAZZVAR dataset, a collection of 502 pairs of Variation and Original MIDI segments.

Abstract

Jazz pianists often uniquely interpret jazz standards. Passages from these interpretations can be viewed as sections of variation. We manually extracted such variations from solo jazz piano performances. The JAZZVAR dataset is a collection of 502 pairs of Variation and Original MIDI segments. Each Variation in the dataset is accompanied by a corresponding Original segment containing the melody and chords from the original jazz standard. Our approach differs from many existing jazz datasets in the music information retrieval (MIR) community, which often focus on improvisation sections within jazz performances. In this paper, we outline the curation process for obtaining and sorting the repertoire, the pipeline for creating the Original and Variation pairs, and our analysis of the dataset. We also introduce a new generative music task, Music Overpainting, and present a baseline Transformer model trained on the JAZZVAR dataset for this task. Other potential applications of our dataset include expressive performance analysis and performer identification.

JAZZVAR: A Dataset of Variations found within Solo Piano Performances of Jazz Standards for Music Overpainting

TL;DR

This paper outlines the curation process for obtaining and sorting the repertoire, the pipeline for creating the Original and Variation pairs, and the analysis of the JAZZVAR dataset, a collection of 502 pairs of Variation and Original MIDI segments.

Abstract

Jazz pianists often uniquely interpret jazz standards. Passages from these interpretations can be viewed as sections of variation. We manually extracted such variations from solo jazz piano performances. The JAZZVAR dataset is a collection of 502 pairs of Variation and Original MIDI segments. Each Variation in the dataset is accompanied by a corresponding Original segment containing the melody and chords from the original jazz standard. Our approach differs from many existing jazz datasets in the music information retrieval (MIR) community, which often focus on improvisation sections within jazz performances. In this paper, we outline the curation process for obtaining and sorting the repertoire, the pipeline for creating the Original and Variation pairs, and our analysis of the dataset. We also introduce a new generative music task, Music Overpainting, and present a baseline Transformer model trained on the JAZZVAR dataset for this task. Other potential applications of our dataset include expressive performance analysis and performer identification.
Paper Structure (24 sections, 2 equations, 5 figures, 4 tables)

This paper contains 24 sections, 2 equations, 5 figures, 4 tables.

Table of Contents

  1. Introduction
  2. Related Works
  3. JAZZVAR Dataset
  4. Data Collection
  5. Repertoire
  6. Jazz Standard Lead Sheets
  7. Audio of Jazz Solo Piano Performances
  8. Automatic Music Transcription of Jazz Audio
  9. Pair Matching Process
  10. Analysis
  11. Experimental Dataset Analysis
  12. Pitch Class Entropy
  13. Pitch Range
  14. Polyphony
  15. Number of Pitches
  16. ...and 9 more sections

Figures (5)

  • Figure 1: The process of creating Original and Variation pairs. Original sections are MIDI segments from a lead sheet transcription of a jazz standard. Audio of a piano performance playing the same jazz standard is transcribed automatically into MIDI. A Variation is found by manually searching for passages that are melodically and harmonically similar to the Original in the "head" section of the piano performance.
  • Figure 2: The melodic contours of the melody taken from the jazz standard "All The Things You Are" (in Blue) and pianists' interpretations.
  • Figure 3: A line graph comparison of the Harmonic Rhythm of the original melody (in Blue) and pianists' interpretations of the melody.
  • Figure 4: Piano-rolls of two Original (left in Blue) and the corresponding generated Variation (right in Red) sections. The Original A is from the song "All the Things You Are", and the Original B is from the song "Alfie".
  • Figure :