Support material and source code for the following work: S.I. Mimilakis, E. Cano, J. Abesser, G. Schuller, "New Sonorities for Jazz Recordings: Separation and Mixing using Deep Neural Networks", in Proceedings of the 2nd AES Workshop on Intelligent Music Production, London, UK, 13 September 2016.
Please use the above citation if you find any of the code useful.
For code usage, please refer to each class. Examples are given inside method or in the "main()" call.
The research leading to these results has received funding from the European Union's H2020 Framework Programme (H2020-MSCA-ITN-2014) under grant agreement no.642685 MacSeNet.