<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META http-equiv=Content-Type content="text/html; charset=gb2312">
<META content="MSHTML 6.00.2900.3268" name=GENERATOR>
<STYLE></STYLE>
</HEAD>
<BODY bgColor=#ffffff>
<DIV><FONT size=2>Hi there, </FONT></DIV>
<DIV><FONT size=2>I like this idea too.</FONT></DIV>
<DIV><FONT size=2>Is it available for GSoC 2008 now? Because I still have
not found the idea in the "SoC ideas" list.</FONT></DIV>
<DIV><FONT size=2>However, I think the hypothesis of making a powerful songs
database/search engine is kind of optimistical now. It's true
that there is abundant information and metadata through Clam
extractor and web services. Nevertheless, I found in state of the art
there are always problems of "semantic gap" of mapping
the extracted low-level information to high-level
descriptors. </FONT></DIV>
<DIV><FONT size=2>But still, we should head for this object.I found there
are several well related ideas such as enhancing chord detection and
annotation merger in the list. </FONT></DIV>
<DIV><FONT size=2></FONT> </DIV>
<DIV><FONT size=2></FONT> </DIV>
<DIV><FONT size=2>Best,</FONT></DIV>
<DIV><FONT size=2>June</FONT></DIV></BODY></HTML>