Skip to content

Latest commit

 

History

History
77 lines (54 loc) · 2.66 KB

README.md

File metadata and controls

77 lines (54 loc) · 2.66 KB

DISCO -- DIStributed semi-COnstrained microphone arrays

This repository gathers scripts to:

Install

In an anaconda environment

To install the content of this package inside an anaconda environment, make sure that:

  1. the environment has been created (see official conda documentation for more info)
  2. the environment is currently active

Then follow the steps outlined in the next subsection.

Auto-install

Type make install from the command line.

Note: The installation of soundfile may fail as it requires libsndfile to be installed on the machine. To make it succeed, make sure libsndfile is installed before running make install.

To generate datasets:

Please refer to the dedicated section

Tests & Documentation

Tests

Tests provide examples of how functions and classes are to be used in addition to ensuring that modifications made to the source code are backward compatible. The test files are all under the directory named tests.

To run the tests, simply type pytest from the command line.

To generate a coverage report, type make coverage from the command line. This will generate a directory named coverage_html_report, which contains the coverage report.

Documentation

To generate the documentation, type make doc from the command line.

Shortcut

In case one wishes to generate a coverage report as well as the documentation, one simply needs to type make from the command line. Under the hood, the commands make coverage and make doc will be run.

References

[1] DNN-based mask estimation for distributed speech enhancement in spatially unconstrained microphone arrays

@inproceedings{Furnon2020SE,
    title={DNN-based mask estimation for distributed speech enhancement in spatially unconstrained microphone arrays},
    author={Furnon, Nicolas and Serizel, Romain and Illina, Irina and Essid, Slim},
    year={2020},
    booktitle={submitted to TASLP},
}

[2] Distributed speech separation in spatially unconstrained microphone arrays

@inproceedings{Furnon2020SS,
    title={Distributed speech separation in spatially unconstrained microphone arrays},
    author={Furnon, Nicolas and Serizel, Romain and Illina, Irina and Essid, Slim},
    year={2020},
    booktitle={submitted to ICASSP},
}