Separating vocals and instruments from music clips with deep learning models
Remember to add project root absolute path (top-level directory of this project) to environment variable $PYTHONPATH
source pythonpath.sh
python3 fetch_dsd100.py
the DSD100.zip will be downloaded in data directory and unzip
python3 generate_dataset.py
options: -t [stft, logstft]
default is STFT transform