Subject: Call for participation: 2nd Community-based Signal Separation Evaluation Campaign (SiSEC 2010) From: Emmanuel Vincent <emmanuel.vincent@xxxxxxxx> Date: Sat, 6 Feb 2010 16:33:58 +0100 List-Archive:<http://lists.mcgill.ca/scripts/wa.exe?LIST=AUDITORY>Dear list, I am posting this announcement on behalf of the organizers of the SiSEC 2010 signal separation evaluation campaign. This campaign is definitely of interest to those of you working on Computational Auditory Scene Analysis. The variety of datasets has increased a lot since SiSEC 2008, with new datasets involving e.g. moving sources or real-world diffuse noise. Looking forward to your entry, Emmanuel -- Emmanuel Vincent METISS Project-Team INRIA Rennes - Bretagne Atlantique Campus de Beaulieu, 35042 Rennes cedex, France Phone: +332 9984 2269 - Fax: +332 9984 7171 Web: http://www.irisa.fr/metiss/members/evincent/ *** CALL FOR PARTICIPATION *** (apologies for cross-posting) 2nd Community-based Signal Separation Evaluation Campaign (SiSEC 2010) http://sisec.wiki.irisa.fr/ Submission deadline: April 16, 2010 SiSEC2010 is the second community-based endeavor for the evaluation of source separation algorithms. Following the first call, seven Audio datasets & three Biomedical/BCI datasets have been proposed by researchers: * Audio source separation 1: Underdetermined speech and music mixtures - Determined and over-determined speech and music mixtures 2: Determined and over-determined speech and music mixtures 3: Robust blind linear/non-linear separation of short two-sources-two-microphones recordings 4: Overdetermined speech and music mixtures for human-robot interaction 5: Determined convolutive mixtures under dynamic conditions 6: Professionally produced music recordings 7: Source separation in the presence of real-world background noise * fMRI or/and microarray data separation 8: Cancer Pathway Reconstruction using Breast Cancer Microarray gene expression profiles * EEG data extraction 9: Dependent component extraction 10: Artefact removal in EEG data Help make this evaluation the most complete yet by trying your source separation algorithms over one or more datasets and submitting the results. To download the datasets and view submission guidelines and evaluation procedures, see http://sisec.wiki.irisa.fr/. Interested individuals are encouraged to join the SiSEC mailing list at https://listes.irisa.fr/sympa/info/sisec. Announcements (including on the revision to the datasets, reference software, submission information, etc.) and discussions related to the SiSEC2010 will be provided through the SiSEC mailing list. Best regards, SiSEC2010 Evaluation Chairs Shoko ARAKI, Fabian Theis and Guido Nolte sisec-org@xxxxxxxx