Music Source Separation with Generative Adversarial Network and Waveform Averaging

Ryosuke Tanabe, Yuto Ichikawa, Takanori Fujisawa, Masaaki Ikehara

研究成果: Conference contribution

抄録

The task of music source separation is to extract a target sound from mixed sound. A popular approach for this task uses a DNN which learns the relationship of the spectrum of mixed sound and one of separated sound. However, many DNN algorithms does not consider the clearness of the output sound, this tends to produce artifact in the output spectrum. We adopt a generative adversarial network (GAN) to improve the clearness of the separated sound. In addition, we propose data augmentation by pitch-shift. The performance of DNN strongly depends on the quantity of the dataset for train. In other words, the limited kinds of the training datasets gives poor knowledge for the unknown sound sources. Learning the pitch-shifted signal can compensate the kinds of training set and makes the network robust to estimate the sound spectrum with various pitches. Furthermore, we process the pitch-shifted signals and average them to reduce artifacts. This proposal is based on the idea that network once learned can also separate pitch-shifted sound sources not only original one. Compared with the conventional method, our method achieves to obtain well-separated signal with smaller artifacts.

本文言語English
ホスト出版物のタイトルConference Record - 53rd Asilomar Conference on Circuits, Systems and Computers, ACSSC 2019
編集者Michael B. Matthews
出版社IEEE Computer Society
ページ1796-1800
ページ数5
ISBN(電子版)9781728143002
DOI
出版ステータスPublished - 2019 11
イベント53rd Asilomar Conference on Circuits, Systems and Computers, ACSSC 2019 - Pacific Grove, United States
継続期間: 2019 11 32019 11 6

出版物シリーズ

名前Conference Record - Asilomar Conference on Signals, Systems and Computers
2019-November
ISSN(印刷版)1058-6393

Conference

Conference53rd Asilomar Conference on Circuits, Systems and Computers, ACSSC 2019
CountryUnited States
CityPacific Grove
Period19/11/319/11/6

ASJC Scopus subject areas

  • Signal Processing
  • Computer Networks and Communications

フィンガープリント 「Music Source Separation with Generative Adversarial Network and Waveform Averaging」の研究トピックを掘り下げます。これらがまとまってユニークなフィンガープリントを構成します。

引用スタイル