Music sample identification dataset enables better matching of transformed tracks
Building a Dataset for Music Sample Identification
Sound
Summary
Identifying music samples means recognizing parts of one song reused in others after being changed. Because previous datasets were too small, the authors created a much larger collection of songs with known sampling information. They carefully split this data to avoid mixing related tracks in both training and test sets, making sure their tests are fair. This new dataset can help researchers create better tools to find music samples.
What this means in practice
- •For music streaming services: Improve detection of sampled audio across large music libraries to enhance recommendations and copyright management.
- •For music producers: Identify source samples used in tracks to aid in creative sampling and legal clearance processes.
Authors
R. Oguz Araz, Xavier Lizarraga, Xavier Serra, Dmitry Bogdanov
Abstract
Sample identification (SI) is the task of matching an element of a musical work to its musically transformed versions used to create new works. The task has received little attention and lacks large-scale publicly available data. In this work, we mine sampling annotations from a music database and split them for training and evaluation. The resulting dataset is nearly three orders of magnitude larger than the existing SI benchmarks, with training, validation, and test sets of 114 k, 6 k, and 10 k tracks. We find that naively splitting the annotations places the same tracks in different sets. To avoid this, we construct a graph from the annotations and split it over connected components. We further find that a single mega-component contains half of the annotations, making component-wise splitting incompatible with balanced splits; we trim it, yielding a leakage-aware pipeline. We share the dataset for non-commercial scientific research purposes only and make the data-analysis and splitting code publicly available. We hope that our work fosters research on SI.