.png){kind=icon}
返回
Music Source Separation With Generative Flow
.png){kind=icon}
OA
AI
DOI：10.1109/LSP.2022.3219355.png)
.png){kind=icon}
原文求助.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
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.
Keyword:
Generative source separation
glow
singing voice separation
music source separation
期刊
IF:
9.6
论文数:
1.1W
被引数:
1.7W
机构
引用论文
Foliar quality of co-occurring mallee eucalypts: balance of primary and secondary metabolites reflects past growing conditions共存桉树小乔木的叶片质量：初级和次级代谢物的平衡反映了过去的生长条件.png)
Chemoecology
ChemoecologyIF0
PREAI
分享收藏分享收藏Correction: Karyotype evolution in Phalaris (Poaceae): The role of reductional dysploidy, polyploidy and chromosome alteration in a wide-spread and diverse genus更正：Phalaris（禾本科）的核型进化：减数异倍性、多倍化和染色体变异在一个广泛分布且多样化的属中的作用PLOS ONE
OAAI
PLOS ONEIF0
OAAI分享收藏Age, Growth, and Reproductive Biology of Three Catostomids From the Apalachicola River, Florida
PREAI
分享收藏分享收藏分享收藏
查看更多内容