[Paper (arXiv)] https://arxiv.org/pdf/2008.07142.pdf

[Github] https://github.com/music-x-lab/POP909-Dataset

A dataset contains multiple versions of the piano arrangements of 909 popular songs created by professional musicians.

The main body of the dataset contains the vocal melody, the lead instrument melody, and the piano accompaniment for each song in MIDI format, which are aligned to the original audio files.

Annotations of tempo, beat, key, and chords. Here, the tempo curves are hand-labeled and others are done by MIR algorithms

Each piano arrangement is stored in MIDI format with three tracks. Figure 2 shows an example of a three-track MIDI file, in which different tracks are labeled with different colors. The three tracks are:

• MELODY: the lead (vocal) melody transcription.
• BRIDGE: the arrangement of secondary melodies or lead instruments.
• PIANO: the arrangement of the main body of the accompaniment, including broken chords, arpeggios, and many other textures.

How to work with Dataset?

1) Fork the git to your repository, and copy the URL

2) type git clone [URL] on your server: It will automatically clone all the content of your forked repository.