« XTech Day 3 - Rufus Pollock and Jo Walsh talk about 'Atomisation and Open Data' | Main | Synchronicity »
18 May 2007
XTech, Quakr
Posted by Rob Styles at May 18, 2007 12:01 PM
Yesterday I sat through an excellent session from three people, each giving a different aspect of how they've gone about building Quakr
From left, Peter Arbuthnott, Katie Portwin and David Sant have been building a 3D tour of our world using photos of Flickr, geo-tagging, and some very expensive proprietary hardware developed by the team.
At the moment they've got some really nice demos showing Quakr's potential, these are centered on Oxford; their home town.
Essentially, what the team have done is take photos using their specialised camera, recording seven aspects of positioning data. These are (from their site)
- Altitude (Are we standing on a mountain?)
- Latitude (How for north/south of the equator are we.)
- Longitude (How far east/west of the meridian are we.)
- Compass bearing (ie, N/S/E/W - which direction are we pointing the camera?)
- Tilt (ie, are we pointing it up at the sky a bit, or down at the ground a bit?)
- Orientation (is this photo portrait, landscape, somewhere wacky in between?
- Timestamp (good for knowing if this is day or night)
This allows them to position the photos accurately in 3D space.
Katie's explanations of the issues involved in dis-ambiguating tags, reconciling different definitions of 'tilt' and then working with "Image Jungles" those spots on the map where there is an over-abundance of photos, was incredibly clear and helped me clarify some of the work I'm doing with bibliographic data.
They're encountering the same kinds of problems we have - that metadata being recycled for uses other than the original purpose is hard to handle and often needs a lot a lot of best-guess cleaning.
In the interests of full-disclosure I have to let you know that we thought these guys were pretty cool before seeing them talk. Some of us had met them at last year's Xtech and we'd been out for dinner the night before. I ordered a dozen escargot and was appalled when both Dave and Katie dug in, but my fellow Talisians declined with such base comments as "I don't eat Mollusks".
Seriously though, what Peter, Katie and Dave are doing stacks up against the (much slicker, and substantially better funded) Photosynth from Microsoft Live Labs.
That these guys have done this as a spare time project is ******* awesome.
Trackback Pings
TrackBack URL for this entry:
http://blogs.talis.com/mt/mt-tb.r280.cgi/888
