09/12/2008

Look Up, Not Down!

monkey.png

Sekai Camera, which has been written up quite a bit since launching at TechCrunch50 on Tuesday, is basically a realtime data overlay onto video from the iPhone's camera. While I'm not sure walking around looking "through" an iPhone is an optimal execution, the idea of a localized data / social communication heads-up display is pretty fascinating.

It was unclear from the unique presentation and Q & A how "real" the system is. The language barrier and clear enthusiasm of the company's executives has reduced my confidence in my own comprehension to approximately 50%, but it seems the current version is based exclusively on GPS location data. The demo video implies an object recognition component, which seems potentially feasible (if exceptionally difficult) in the very near future. The iPhone does have the ability to stream live video, and recent demonstrations have shown the feasibility of realtime remote rendering and other huge computational tasks.

The panel at TC50 raised other important questions, like "how is data kept current"? Seeing firsthand the wonder that was TonchiDot's presentation, my personal guess is that it may be a while before we get some answers.

Christian Dodd

TrackBack

TrackBack URL for this entry:
http://threeminds.organic.com/cgi-bin/movabletype/mt-tb.cgi/6910

Comments (4)

mike:

Seems like this will break down along topical areas...like getting more stats for a player while watching sports, or something like the museum tour they show in the vid. one-size-fits all just kind of fries my mind thinking about.

it seems like most of the answers this thing is providing could be achieved with a google search from your phone (or from the printed labels and signs nearby, gasp.) that's a cool interface though. cute as a button. is there more to their current product than the interface?

Patten Author Profile Page:

The implications of a device like this are great and far-reaching. There is no doubt that the current technical feasibility for the SekaiCamera is quite small - in addition to the bandwidth required to get this type of data to your mobile device in a truly real-time fashion, there is also the issue of the mobile device's ability to process all that data and display it, not to mention of the device's camera which would need to have some serious optical recognition.

Still, despite these shortcomings, it would be an incredibly powerful and useful technology. Let me explain how with a recent story.

My friend came in from Texas this past weekend to hang out with me but, more importantly, to go visit Yankee Stadium. He has always been a Yankees fan but had never been to the stadium. Now that it was set to be demolished, it was imperative for him to not only see the stadium but also to make it to Monument Park and take in as much of the surroundings as he could. So, we traveled early Saturday morning on the 4 train to the Bronx and proceeded to check out as much as we could. The thing that he noticed was that there was not a lot of signage in the park and there were very few things about the stadium itself that instantly jumped out at you as historical landmarks (e.g. there was no “Green Monster” like at Fenway). However, I assume that many of the aspects of the stadium have exceptional historical value. Doing a little digging on Wikipedia when we got back to my apartment later that night, we found that the park does indeed have many distinguishing characteristics. However, none of these are instantly available for interrogation at the stadium itself and, at a minimum, likely require purchasing a program where we would’ve had to sift through the 8,291 ads contained therein. Having a device like the SekaiCamera would’ve meant we could’ve just sat in our (very good thanks to a friend) seats and allowed the device to do the interrogation for us. It would’ve made the entire experience that much more enjoyable and informative, let alone memorable, for both of us.

Marta Strickland Author Profile Page:

I can totally relate to your story, Chris... when my husband and I were visiting the Imperial Forums in Rome, I could not believe there were NO signs at all to describe what I was looking at. I mean here were thousand year old ruins, one of which had a tile floor. My husband and I had no way of knowing whether those tiles were restored, recreated, or just an artistic choice of someone many years later.

The usefulness is definitely easy to grasp. But, I wonder about the implementation. Couldn't the solution just be a smarter interface for Google Earth community notations available via your mobile phone? It seems like a technology we are so very close to already, without the need for this far-reaching idea of streaming image recognition and augmented reality.

Sebastien Monney:

While some of these concepts may feel far-fetched...Japan is more comfortable w/mobile technology than we are (have you purchased anything with your phone yet?), and their 3G network probably works as advertised ;)
So yeh, it's possible, look up!

Post a comment


Type the characters you see in the picture above.