Miron M, Janer J, Gomez E. Generating data to train convolutional neural networks for low latency classical music source separation. Proceedings of the 14th Sound and Music Computing Conference
List of results published directly linked with the projects co-funded by the Spanish Ministry of Economy and Competitiveness under the María de Maeztu Units of Excellence Program (MDM-2015-0502).
List of publications acknowledging the funding in Scopus.
The record for each publication will include access to postprints (following the Open Access policy of the program), as well as datasets and software used. Ongoing work with UPF Library and Informatics will improve the interface and automation of the retrieval of this information soon.
The MdM Strategic Research Program has its own community in Zenodo for material available in this repository 
as well as at the UPF e-repository 
Miron M, Janer J, Gomez E. Generating data to train convolutional neural networks for low latency classical music source separation. Proceedings of the 14th Sound and Music Computing Conference
Miron M, Janer J, Gomez E. Generating data to train convolutional neural networks for low latency classical music source separation. Proceedings of the 14th Sound and Music Computing Conference
Deep learning approaches have become increasingly popular in estimating time-frequency masks for audio source separation. However, training neural networks usually requires a considerable amount of data. Music data is scarce, particularly for the task of classical music source separation, where we need multi-track recordings with isolated instruments. In this work, we depart from the assumption that all the renditions of a piece are based on the same musical score, and we can generate multiple renditions of the score by synthesizing it with different performance properties, e.g. tempo, dynamics, timbre and local timing variations. We then use this data to train a convolutional neural network (CNN) which can separate with low latency all the renditions of a score or a set of scores. The trained model is tested on real life recordings and is able to effectively separate the corresponding sources. This work follows the principle of research reproducibility, providing related data and code, and can be extended to separate other pieces.
Additional material:
- Post-print at UPF repository
- Slides
- Code available on the source separation github repository DeepConvSep