PhotoSynth: What. How. Why.

Posted By: Charles | Jul 28th, 2006 @ 10:30 AM | 142,243 Views | 85 Comments

Formats:

Blaise Aguera y Arcas is an Architect of PhotoSynth, which is a super cool 3-D image "tourism" application that enables a new methodology for exploring related groups of images using a complex imaging algorithm developed in part by Microsoft Research. Here, Charles sits down with Blaise to learn the details of the What How and Why of this highly innovative distributed 3-D image processing application. PhotoSynth enables a whole new way of exlporing images! Wow.

To find out more on how to use Photosynth, watch this video on 10.

Tags: MS Research, PhotoSynth, Windows Live

Rating:

Chadk
excuse me - do you has a flavor?

WOW! I have only seen 7 minuts.
But the demo 7 minuts into the movie is AMAZING!!

jsampsonPC
SampsonBlog.com SampsonVideos.com

Scoble laugh at 00:00:54 Smiley

Scoble's spirit lives on in Charles.

Charles
Welcome Change

This technology is outrageously cool. Listen carefully to what Blaise talks about in regards to distributed imaging over the network. Imagine navigating through a 3-D image composed of a collection of images created by thousands of people all over the world. The implications here are gigantic. I'm still interpreting what I learned. It's amazing.

C

jsampsonPC
SampsonBlog.com SampsonVideos.com

Charles, I'm still a little confused as to the first part. He stated that as the images are navigated (zooming in and out), they are simulatenously being opened and closed? Is he implying that each picture has several versions of differing quality? I don't think he is, because that would see ridiculous to store 20 versions of 20 differign qualities for every 1 image on your PC. Could you please explain what is taking place behind the scenes as the user navigates the canvas of canvas's? Smiley

hold on, does it have to do with the "multi-resolution"?

HumanCompiler
Compiling humans...and code

I'd love to use this using Play Anywhere!

What I think is the best part of all this is not the ideas in general. Breaking images into pieces (like a pyramid) to zoom in and out of images efficiently or stiching photos together to give a 3D like experience have been around for a long time. 2 jobs ago we used a fish-eye lens and ViewPoint to do this. WHAT IS MOST AMAZING (to me anyway) about this stuff is that what it generates is in 3D space, but most importantly any joe schmoe can take pictures with their digital camera and use this. That takes the audience for this type of thing from not very many to like half the planet (just a guess). This is huge! Then take the distributed imaging that Charles mentioned on top of that and the implications for what all this could affect is mind blowing.

Charles
Welcome Change

I'd prefer Blaise to answer your questions here Smiley I am far from qualified to explain this stuff...

C

jsampsonPC
SampsonBlog.com SampsonVideos.com

Is Blaise online as a Niner?

I am still wondering how exactly it shows "versions?" of the images as you zoom in and out, if it's drawing from a single image...does it only display a portion of the image's binary?

XX00XX00XX00
00XX00XX00XX
XX00XX00XX00
00XX00XX00XX
XX00XX00XX00
00XX00XX00XX

Displayed with a Z value of .5 is reduced to

X0X0X0
0X0X0X
X0X0X0
0X0X0X

And then at a Z value of .25 it is

XXX
000

Or something similar? No, no no....It couldn't be this way - you would still have to load the entire binary source to even do this....argh, I'm too stupid to be a niner Smiley

Baise, if you're here - how exactly does this "Multi-resolution" aspect of the image's binary work?

jsampsonPC
SampsonBlog.com SampsonVideos.com

IDEA!
Blaise, or Charles: Has this technology ever been considered in Digital Microscopy?

Charles
Welcome Change

Microscopy is certainly an area where this could be highly useful. Consider also astronomy... Navigating galaxies and other celestial bodies will never be the same! Smiley