Skip to content
/ mirapie Public
forked from Chutlhu/mirapie

Interference removal algorithm for multitrack live recordings

License

Notifications You must be signed in to change notification settings

TGabor/mirapie

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

19 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

mirapie

Multitrack Interference RemovAl for full-length live recordings

Example:

Using the provided toydata:

$./mirapie.py  toydata/ toydata/initL.csv -p 1 -m 1

Using a your multitrack dataset, e.g. in ../path/to/dataset and the interference matrix in ../path/to/matrix.csv.

$./mirapie.py    ../path/to/dataset/   ../path/to/matrix.csv -p 1 -m 3

Installation

1 Setup the virtual enviroment, run it and install all the dependencies

$ virtualenv  venv -p python3
$ source venv/bin/activate
$ pip install -r requirements

Usage

$ ./mirapie.py -h
    MIRAPIE: Multitrack Interference RemovAl for full-lenght live recordings
    usage: 
        mirapie.py  path-to-wavs csv-matrix [-h] [-l MATRIX] [-p PRESET] [-m MODE]
    Python implementation for MIRA
    positional arguments:
        path-to-wavs        location of the multitrack recordings [.wav]
        csv-matrix          name of the initial interference matrix [.csv]
    optional arguments:
    -h, --help              show this help message and exit
    -l MATRIX, --matrix MATRIX   
                            L matrix file (default: None)
    -p PRESET, --preset PRESET
                            select one of the possible preset (default: 1)
    -m MODE, --mode MODE    select one of the possible mode (default: 0)

Command line arguments

path-to-wavs : folder containing the audio recording in [.wav]
csv-matrix : file containing info about mic channels and instrumens organized in a table [.csv]
-p PRESET : present number in the user-editable yaml file preset.yml
-m MODE : mode of the algorithm, 1 one chunk, 2 one chunk with random projection, 3 full-length with random projection.

advance algorithm parameters in preset.yml file

Author

Diego Di Carlo (Chutlhu)

Notes

  • Works only with Python3
  • Conference paper about mirapie can be downloaded here
  • Special thanks to Antoine Liutkus and Thomas Praetzlich and Multispeech Team at INRIA GRAND EST.

Development

Troubleshooting

License

GPL v3

About

Interference removal algorithm for multitrack live recordings

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 100.0%