dsdtools

A python package to parse and process the demixing secrets dataset (DSD) as part of the MUS task of the Signal Separation Evaluation Campaign (SISEC)

Contents:

• Installation
  • DSD100 Dataset / Subset
• Usage
  • Providing a compatible function
  • Create estimates for SiSEC evaluation
    • Setting up dsdtools
    • Test if your separation function generates valid output
    • Processing the full DSD100
    • Processing training and testing subsets separately
    • Processing single or multiple DSD100 tracks
    • Use multiple cores
  • Evaluation in python
• Example
• Modules
  • Modules
  • Audio Classes

References

If you use this package, please reference the following paper

@inproceedings{SiSEC2015,
  TITLE = {{The 2015 Signal Separation Evaluation Campaign}},
  AUTHOR = {N. Ono and Z. Rafii and D. Kitamura and N. Ito and A. Liutkus},
  BOOKTITLE = {{International Conference on Latent Variable Analysis and Signal Separation  (LVA/ICA)}},
  ADDRESS = {Liberec, France},
  SERIES = {Latent Variable Analysis and Signal Separation},
  VOLUME = {9237},
  PAGES = {387-395},
  YEAR = {2015},
  MONTH = Aug,
}