Special Sessions

Following the increased interest for real-world applications of LVA and BSS, we are proposing to hold two special sessions, giving you the opportunity to experience the latest exciting developments in this field: 

1)   Real-world constraints and opportunities in audio source separation, including:

• real-time, online, or computationally efficient algorithms
• use of extra available information (musical score, etc)
• reduction of sparkling artifacts (aka musical noise)
• FGPA/hardware implementation
• benchmarking of real-world implementations 

2)   From audio source separation to multisource content analysis, including:

• interfacing BSS or speech enhancement with ASR,AED or MIR
• using NMF, NTF or other latent variable models for ASR, AED or MIR 

You may participate by presenting either:

• an original research paper (to be reviewed by the organizers and the other authors of the session, and to be included in the conference proceedings, published at Springer's LNCS)
• a poster or a demo of a previously published or a recent result (all such proposals will be accepted in the limit of available space, but will not be included in the conference proceedings)

Full research papers and/or 10-line abstracts of posters/demos must be submitted by Tuesday, December 6^th, 2011.

As usual, participants will be expected to register to the conference and pay registration fees upon acceptance.

If you are willing to participate in either of these sessions, please send an email to: 

• your name and affiliation
• the title and type (paper/poster/demo) of your contribution
• whether you would have attended LVA/ICA anyway without these special sessions (this information is needed to improve our service to the conference audience)