jump to navigation

Seadragon June 18, 2007

Posted by Andre Vellino in Data Mining, Information retrieval, Visualization.
trackback

Have a look at this 8-minute video of Seadragon, the Microsoft Labs application which demonstrates how a 3-d model of Notre Dame can be induced from a large number of photo-stills (thanks to Glen for pointing this out.) Some aspects of the demo are based on “Photosynth“, which you can try on your PC (DirectX-based, I believe.)

Now, imagine a generalization of this to the world of scientific knowledge that could be induced from scholarly books and articles. In fact, you could see that there may be large numbers of ways in which documents could be “mapped” in this way. The advantage of a building is that it’s fixed in space and with a large collection of snapshots you can iterate to a “fixed point”. With concepts and words this may be a little more difficult….

OK, maybe that’s thinking too far ahead, but one can at least imagine all kinds of nested, next-generation “hyperlinks” from documents, to data, to images (e.g. of micro-organisms or stellar objects) to computer models (proteins, large molecules).

Comments»

1. Peter Turney - June 19, 2007

Now, imagine a generalization of this to the world of scientific knowledge that could be induced from scholarly books and articles.

This is a nice idea. The closest approximation that comes to mind is some of Oren Etzioni’s work:

Open Information Extraction from the Web
http://turing.cs.washington.edu/papers/ijcai07.pdf

Relational Web Search
http://www.cs.washington.edu/homes/etzioni/papers/UW-CSE-06-04-02.pdf

Unsupervised Resolution of Objects and Relations on the Web
http://turing.cs.washington.edu/papers/object_identification_camera_ready_4.pdf

Oren Etzioni
http://www.cs.washington.edu/homes/etzioni/

2. Is Everything Miscellaneous? « Synthèse - June 22, 2007

[…] for example the impressive Seadragon demo that reverse-engineers a 3-D model of Notre Dame from photographs. If everything is miscellaneous, […]

3. alan - January 21, 2011

how do we use the software that the presentor used for the 3d model of notre dame

Andre Vellino - January 30, 2011

I think this software was experimental and developed at the University of Washington.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: