Sound processing & MIR - any tips?

[Marcin]

Dear Participants,

ISMIS 2011 Contest is a scientific event. The goal is not only to solve the problem and win awards, but - more importantly - to learn about the fascinating fields of sound recognition and Music Information Retrieval, because many of us, data miners, don't know much about it, yet. So, all of you who are more familiar with these domains and can share any useful information, hints, links to online materials and articles, explain the meaning of sound features in the data ... - feel free to do this! Give all of us a chance to learn something new!

You may post in this thread. Posts can be loosely related to the contest: put just any links and information related to sound processing and MIR that you think might be interesting for others. Discussions are welcome.

Thanks!

Marcin Wojnarski
The Organizers

[artificial]

I found some articles that might be useful, check this out:
1. AUTOREGRESSIVE MFCC MODELS FOR GENRE CLASSIFICATION IMPROVED BY HARMONIC-PERCUSSION SEPARATION
2. MODIFIED AIS-BASED CLASSIFIER FOR MUSIC GENRE CLASSIFICATION
3. A PROBABILISTIC SUBSPACE MODEL FOR MULTI-INSTRUMENT POLYPHONIC TRANSCRIPTION

Let me know if that helped...........