/deep-music-enhancer

Primary LanguagePythonOtherNOASSERTION

Source code for paper "On Filter Generalization for Music Bandwidth Extension Using Deep Neural Networks", Serkan Sulun, Matthew E. P. Davies, 2020. https://arxiv.org/abs/2011.07274v2

To cite:

S. Sulun and M. E. P. Davies, "On Filter Generalization for Music Bandwidth Extension Using Deep Neural Networks," in IEEE Journal of Selected Topics in Signal Processing, vol. 15, no. 1, pp. 132-142, Jan. 2021, doi: 10.1109/JSTSP.2020.3037485.

Required Python libraries: Numpy, Scipy, Pytorch, Requests, tqdm

For CUDA 11.0 you can also run pip install -r requirements.txt

To download datasets: Run get_datasets.py within folder datasets. Only downloads DSD100, since MedleyDB requires special permission.

To download pre-trained models: https://drive.google.com/drive/folders/19hltS3bSJFdXkxILlAC_OaeRSwKZn2Cc?usp=sharing

config.py specifies hyperparameters, command-line arguments and general configuration.

Main script: run.py within folder src

Tested with Python 3.7.9, Numpy 1.19.4, Scipy 1.6.0, Pytorch 1.7.1

Acknowledgements

Serkan Sulun receives the support of a fellowship from ”la Caixa” Foundation (ID 100010434), with the fellowship code LCF/BQ/DI19/11730032. This work is also funded by national funds through the FCT - Foundation for Science and Technology, I.P., within the scope of the project CISUC - UID/CEC/00326/2020 and by European Social Fund, through the Regional Operational Program Centro 2020 as well as by Portuguese National Funds through the FCT - Foundation for Science and Technology, I.P., under the project IF/01566/2015.