sigsep-mus-2018 icon indicating copy to clipboard operation
sigsep-mus-2018 copied to clipboard

SiSEC MUS 2018 Submission System

SISEC MUS 2018 Submissions

This is the submission system for the Professionally-produced music task of the 2018 SiSEC.

For the this year, submissions were handled via GitHub where each participant has submitted their results scores using pull requests.

Submissions

Submission Name Blind Add. train. Data
2DFT :heavy_check_mark:
HEL1
HPLP :heavy_check_mark:
HPSS :heavy_check_mark:
IBM1 Oracle
IBM2 Oracle
IRM1 Oracle
IRM2 Oracle
JY1
JY2
JY3
MDL1
MDLT
MELO :heavy_check_mark:
MIX Oracle
MWF Oracle
REP1 :heavy_check_mark:
REP2 :heavy_check_mark:
RGT1
RGT2
RPCA :heavy_check_mark:
STL1 :heavy_check_mark:
STL2
TAK1
TAK2 :heavy_check_mark:
TAK3 :heavy_check_mark:
TAU1 :heavy_check_mark:
UHL1
UHL2
UHL3 :heavy_check_mark:
WK

Analysis

Please refer to the analysis repository.

Paper

to be announced.

Citation

@inproceedings{
  sisec2018,
  title={The 2018 signal separation evaluation campaign},
  author={F.-R. St{"o}ter and A. Liutkus and N. Ito},
  booktitle={International Conference on Latent Variable Analysis and Signal Separation},
  year={2018}
}

Acknowledgements

We thank Colin Raffel for the inspiration to use git for handling the SiSEC submissions.