arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1807.02710cs.SDcs.LGeess.AS
收藏

利用相位特征改进基于DNN的音乐声源分离

Improving DNN-based Music Source Separation using Phase Features

École Polytechnique Fédérale de Lausanne (EPFL)(洛桑联邦理工学院);Sony European Technology Center (EuTEC)(索尼欧洲技术中心);Swiss Data Science Center(瑞士数据科学中心);ETH Zürich(苏黎世联邦理工学院);Sony Corporation(索尼公司)

Joachim Muth, Stefan Uhlich, Nathanael Perraudin, Thomas Kemp, Fabien Cardinaux, Yuki Mitsufuji

摘要:

使用深度神经网络的音乐声源分离通常仅依赖于幅度特征。在本文中,我们表明,额外的相位功能可以提高分离性能。使用STFT相位和幅度之间的理论关系,我们推测,相位的衍生物是一个很好的功能表示反对的原始相位。我们通过实验验证了这一猜想,并提出了一种新的DNN架构,它结合了幅度和相位。与仅使用幅度特征的网络相比,这种联合方法在所有仪器的DSD 100数据集上实现了更好的信号失真比。特别地,低音乐器受益于相位信息。

英文摘要:

Music source separation with deep neural networks typically relies only on amplitude features. In this paper we show that additional phase features can improve the separation performance. Using the theoretical relationship between STFT phase and amplitude, we conjecture that derivatives of the phase are a good feature representation opposed to the raw phase. We verify this conjecture experimentally and propose a new DNN architecture which combines amplitude and phase. This joint approach achieves a better signal-to distortion ratio on the DSD100 dataset for all instruments compared to a network that uses only amplitude features. Especially, the bass instrument benefits from the phase information.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑