Skip to content

A MATLAB implementation of “Multiple Sound Source Counting and Localization Based on TF-Wise Spatial Spectrum Clustering” [TASLP 2019]

License

Notifications You must be signed in to change notification settings

BingYang-20/TF-Wise-Spatial-Spectrum-Clustering

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

12 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

TF-Wise-Spatial-Spectrum-Clustering

A MATLAB implementation of “Multiple Sound Source Counting and Localization Based on TF-Wise Spatial Spectrum Clustering”, IEEE/ACM Transactions on Audio, Speech, and Language Processing (TASLP), 2019.

Main Description

  • MSSL.py is the main implementation of the proposed method
    • SinSouTF.py provides the binary TF weight for single source dominated TF bins and the ranked eigenvectors of spatial correlation matrix
    • TFSpatSpect.py calculates the TF-wise spatilal spectrum (see method description in Section III)
    • SouCouLoc.py joint counts and localizes multiple sound sources (see method description in Section IV)
    • sv.mat stores the steering vector for the considered 8-channel uniform circular microphone array
  • example.py gives an example for multiple sound source localization on the data x.mat
  • x.mat stores one instance of microphone signals (1s, 16000samples, 8 channels)

Quick Start

  • run example.py
  • If you want to change test data, you can change the loaded mat data (currently x.mat) in example.py
  • If you want to change the topolopy of microphone array (number of mircophones or geometry), you need to change the steering vector (currently sv.mat) in MSSL.py according to array topology

Citation

If you find our work useful in your research, please consider citing:

@InProceedings{yang2019TFSSC,
    Author = "Bing Yang and Hong Liu and Cheng Pang and Xiaofei Li",
    Title = "Multiple Sound Source Counting and Localization Based on {TF}-Wise Spatial Spectrum Clustering",
    Journal = "{IEEE/ACM} Transactions on Audio, Speech, and Language Processing (TASLP)",
    Volume = "27",	
    Number = "8",
    Pages = "1241-1255",
    Year = "2019"}

Licence

MIT

About

A MATLAB implementation of “Multiple Sound Source Counting and Localization Based on TF-Wise Spatial Spectrum Clustering” [TASLP 2019]

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages