arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2204.09079eess.AScs.SDeess.SP
收藏

音乐源分离与生成流

Music Source Separation with Generative Flow

University of Rochester(罗切斯特大学);Tencent Technology (Shanghai) Co., Ltd.(腾讯科技(上海)有限公司)

Ge Zhu, Jordan Darefsky, Fei Jiang, Anton Selitskiy, Zhiyao Duan

摘要:

用于源分离的全监督模型是在并行混合源数据上训练的,并且是目前最先进的。然而,这样的并行数据通常难以获得,并且使训练模型适应具有新源的混合物是繁琐的。相比之下,仅源监督模型只需要单个源数据进行训练。在本文中,我们首先利用基于流的生成器来训练个人音乐源先验,然后使用这些模型,以及基于可能性的目标,分离音乐混合物。我们表明,在歌声分离和音乐分离任务中,我们提出的方法与完全监督的方法相比具有竞争力。我们还证明了我们可以灵活地添加新类型的源,而完全监督的方法需要重新训练整个模型。

英文摘要:

Fully-supervised models for source separation are trained on parallel mixture-source data and are currently state-of-the-art. However, such parallel data is often difficult to obtain, and it is cumbersome to adapt trained models to mixtures with new sources. Source-only supervised models, in contrast, only require individual source data for training. In this paper, we first leverage flow-based generators to train individual music source priors and then use these models, along with likelihood-based objectives, to separate music mixtures. We show that in singing voice separation and music separation tasks, our proposed method is competitive with a fully-supervised approach. We also demonstrate that we can flexibly add new types of sources, whereas fully-supervised approaches would require retraining of the entire model.

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。

↑