Moisesdb: A dataset for source separation beyond 4-stems
This dataset enables researchers to build and evaluate fine-grained source separation systems for music, addressing a data limitation in the field, though it is incremental as it extends existing datasets.
The authors introduced the MoisesDB dataset to address the lack of data for musical source separation beyond four stems, consisting of 240 tracks from 45 artists across twelve genres with a hierarchical taxonomy. They provided baseline results for separation models at varying granularities (four, five, and six stems) to facilitate evaluation and adoption.
In this paper, we introduce the MoisesDB dataset for musical source separation. It consists of 240 tracks from 45 artists, covering twelve musical genres. For each song, we provide its individual audio sources, organized in a two-level hierarchical taxonomy of stems. This will facilitate building and evaluating fine-grained source separation systems that go beyond the limitation of using four stems (drums, bass, other, and vocals) due to lack of data. To facilitate the adoption of this dataset, we publish an easy-to-use Python library to download, process and use MoisesDB. Alongside a thorough documentation and analysis of the dataset contents, this work provides baseline results for open-source separation models for varying separation granularities (four, five, and six stems), and discuss their results.