Dear Albrecht,<div><br></div><div>there is lots and lots of work on these topics, you just need to look.</div><div>Marco Baroni and others have tried to produce a joint model of picture</div><div>and distributional word similarity (Bruni et al ACL 2012 Distributional Semantics</div>
<div>in Technicolor), Shane Bergsma produced a visual-features model of</div><div>selectional preferences (Bergsma and Goebel RANLP 2011 Using</div><div>Visual Information to predict selectional preference).</div><div><br>
</div><div>Most of this work on finding object features uses a combination of color-space</div><div>distribution and SIFT features (both probably have a decent implementation</div><div>in OpenCV/SimpleCV).</div><div><br></div>
<div>Movies and pictures can probably be handled by color space statistics</div><div>quite easily, as this aspect is manipulated in post-production and subject</div><div>to deliberate stylistic choice by the people making the film or the picture.</div>
<div>(e.g., National Geographic style jungle realism vs. modern scifi movie</div><div>blue-and-orange tints).</div><div><br></div><div>-Yannick</div><div><br><div class="gmail_quote">On Tue, Nov 13, 2012 at 11:31 AM, Albretch Mueller <span dir="ltr"><<a href="mailto:lbrtchx@gmail.com" target="_blank">lbrtchx@gmail.com</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">~<br>
Many forms of string metrics are used in corpora research, but I<br>
don't see anywhere anything about binary/encoded data. This is what I<br>
have in mind:<br>
~<br>
you feed many paintings and the styles/authors are figured out:<br>
Picassos, Matisses, Kansdinkys, ...<br>
~<br>
object detection inside pictures: these are all shoes ...<br>
~<br>
while watching Michela Watkins' "Bitch Pleeze" performance or her Ann<br>
Coulter impersonation, you want to stop the reel at some point and<br>
have her face at this very moment to match other faces in a corpus,<br>
follow the faces' transitions and check it as a multi-modal input with<br>
their speech context and other gestures<br>
~<br>
... stratify all types of face expressions<br>
~<br>
you want to know when a video segment has been repeated exactly or<br>
approximately/similarly, say, the actual Chaplin silent movies in the<br>
views and rehashings in Richard Attenborough's biographical movie<br>
~<br>
Do you know on any research on those topics? google scholar didn't<br>
give me much good leads<br>
~<br>
lbrtchx<br>
<br>
_______________________________________________<br>
UNSUBSCRIBE from this page: <a href="http://mailman.uib.no/options/corpora" target="_blank">http://mailman.uib.no/options/corpora</a><br>
Corpora mailing list<br>
<a href="mailto:Corpora@uib.no">Corpora@uib.no</a><br>
<a href="http://mailman.uib.no/listinfo/corpora" target="_blank">http://mailman.uib.no/listinfo/corpora</a><br>
</blockquote></div><br></div><br clear="all"><div><br></div>-- <br>Dr. Yannick Versley<div><br><div>Sonderforschungsbereich 833</div></div><div>Universität Tübingen</div><div>Nauklerstr. 35</div><div>72074 Tübingen</div>
<div><br></div><div>Tel.: +49-7071-29 77155</div><br>