Open Dataset

This is a dataset for music source separation provided by Daichi Kitamura.
This dataset consists of following music audio files:
  • Four melody parts

  • Various musical instruments in each part
    • Melody 1: Oboe, Trumpet, and Horn
    • Melody 2: Flute, Violin, and Clarinet
    • Midrange: Piano and Harpsichord (Cembalo)
    • Bass: Trombone, Fagotto (Basson), and Cello

  • Three types of PCM-based MIDI synthesizer for each instrumental sound
    • MGS: Microsoft GS Wavetable Synth
    • YMH: YAMAHA MU-1000
    • GPO: Garritan Personal Orchestra 4

  • Two octave ascending scale tones covering all the tones in each melody part

  • This scale sound is provided only for YMH and GPO tones.
Only the use for academic research is allowed.
If you use this dataset and publish manuscripts, please cite the following paper that explains the details of this dataset:
Daichi Kitamura, Hiroshi Saruwatari, Hirokazu Kameoka, Yu Takahashi, Kazunobu Kondo, and Satoshi Nakamura, "Multichannel signal separation combining directional clustering and nonnegative matrix factorization with spectrogram restoration," IEEE/ACM Transactions on Audio, Speech, and Language Processing, vol. 23, no. 4, pp. 654–669, April 2015.
Also, please cite the URL of this web page.

Download file: zip (32 MB)

    Top     Research Interests     Academic Background     Publications     Demonstrations     Dataset     Program Codes     Work History     Awards     Competitive Funds     Contact me     Links