Tuesday, June 12, 2007
TED 2007
if(window.$WebSecurity){window.$WebSecurity.FlashDetection.HandleEmbedCode('08c1dc49-18e6-4b5b-8eda-5dfa23bec8ae', '\x3cembed src\x3d\x22http\x3a\x2f\x2fstatic.videoegg.com\x2fted\x2fflash\x2floader.swf\x22 flashvars\x3d\x22bgColor\x3dFFFFFF\x26amp\x3bfile\x3dhttp\x3a\x2f\x2fstatic.videoegg.com\x2fted\x2fmovies\x2fBLAISEAGUERAYARCAS-2007_high.flv\x26amp\x3bautoPlay\x3dfalse\x26amp\x3bfullscreenURL\x3dhttp\x3a\x2f\x2fstatic.videoegg.com\x2fted\x2fflash\x2ffullscreen.html\x26amp\x3bforcePlay\x3dfalse\x26amp\x3blogo\x3d\x26amp\x3ballowFullscreen\x3dtrue\x22 quality\x3d\x22high\x22 wmode\x3d\x22window\x22 width\x3d\x22432\x22 height\x3d\x22285\x22 align\x3d\x22middle\x22 type\x3d\x22application\x2fx-shockwave-flash\x22 pluginspage\x3d\x22http\x3a\x2f\x2fwww.macromedia.com\x2fgo\x2fgetflashplayer\x22 allowscriptaccess\x3d\x22never\x22 allownetworking\x3d\x22internal\x22 \x2f\x3e');}else{if(window.flObj == null){window.flObj = new Object();}window.flObj['08c1dc49-18e6-4b5b-8eda-5dfa23bec8ae']='\x3cembed src\x3d\x22http\x3a\x2f\x2fstatic.videoegg.com\x2fted\x2fflash\x2floader.swf\x22 flashvars\x3d\x22bgColor\x3dFFFFFF\x26amp\x3bfile\x3dhttp\x3a\x2f\x2fstatic.videoegg.com\x2fted\x2fmovies\x2fBLAISEAGUERAYARCAS-2007_high.flv\x26amp\x3bautoPlay\x3dfalse\x26amp\x3bfullscreenURL\x3dhttp\x3a\x2f\x2fstatic.videoegg.com\x2fted\x2fflash\x2ffullscreen.html\x26amp\x3bforcePlay\x3dfalse\x26amp\x3blogo\x3d\x26amp\x3ballowFullscreen\x3dtrue\x22 quality\x3d\x22high\x22 wmode\x3d\x22window\x22 width\x3d\x22432\x22 height\x3d\x22285\x22 align\x3d\x22middle\x22 type\x3d\x22application\x2fx-shockwave-flash\x22 pluginspage\x3d\x22http\x3a\x2f\x2fwww.macromedia.com\x2fgo\x2fgetflashplayer\x22 allowscriptaccess\x3d\x22never\x22 allownetworking\x3d\x22internal\x22 \x2f\x3e';}
Subscribe to:
Post Comments (Atom)
It's a cool visualization technology that MS bought in 2006. It's called seadragon. The selling point is fast navigation in large amount of images (raster or vector).
ReplyDeleteThe Notro Dame model adds one more level of coolness to it, which is done by Rick Szeliski and his intern. But I suspect that the 3D model based grouping is only limited to large static object that people will tag in their flickr images (a.k.a. landmark buildings)
It's still meta data-based search, not really content-based search.