Untwist is python library for audio source separation. It provides a self-contained object-oriented framework including common source separation algorithms as well as input/output functions, data management utilities and time-frequency transforms.
To install untwist, clone the github repository and run
python setup.py installfrom the root directory. Alternatively, you can use pip:
pip install git+https://github.com/IoSR-Surrey/untwistThese commands will install all dependencies for you and build the required extension modules for improved performance.
A typical usage scenario involves loading an audio file, computing a
time-frequency transform to obtain a spectrogram, and using some of the source
separation algorithms to obtain a time-frequency mask. The mask is then
multiplied with the spectrogram and the result is converted back to the time
domain. See the API documentation for
reference. See the examples folder for some code examples (you will need to
download this audio dataset.)
The library is under development at the Centre for Vision, Speech and Signal Processing, University of Surrey. The source code is published under the MIT license.