MedleyVox: An Evaluation Dataset for Multiple Singing Voices Separation
This work addresses a rarely studied problem in music source separation for researchers, providing a dataset and baseline methods, but it is incremental as it builds on existing separation techniques.
The authors tackled the lack of a benchmark dataset for multiple singing voices separation by introducing MedleyVox, an evaluation dataset with categorized problem definitions, and proposed an improved super-resolution network (iSRNet) that, when jointly trained, achieved comparable performance to ideal time-frequency masks on duet and unison subsets.
Separation of multiple singing voices into each voice is a rarely studied area in music source separation research. The absence of a benchmark dataset has hindered its progress. In this paper, we present an evaluation dataset and provide baseline studies for multiple singing voices separation. First, we introduce MedleyVox, an evaluation dataset for multiple singing voices separation. We specify the problem definition in this dataset by categorizing it into i) unison, ii) duet, iii) main vs. rest, and iv) N-singing separation. Second, to overcome the absence of existing multi-singing datasets for a training purpose, we present a strategy for construction of multiple singing mixtures using various single-singing datasets. Third, we propose the improved super-resolution network (iSRNet), which greatly enhances initial estimates of separation networks. Jointly trained with the Conv-TasNet and the multi-singing mixture construction strategy, the proposed iSRNet achieved comparable performance to ideal time-frequency masks on duet and unison subsets of MedleyVox. Audio samples, the dataset, and codes are available on our website (https://github.com/jeonchangbin49/MedleyVox).