Google Visual Search: AR 1.5 and Beyond

With the debut of Google Goggles/Visual Search we've seen the immediate (and maybe long-term) future of augmented reality (AR). The apps that show you restaurant reviews floating in space or Tweets in mid air are novel and fun yet ultimately not very useful. As I previously wrote, "augmented reality 1.0" is what we have now.

What Google demonstrated at its Search Evolution event yesterday morning amounts to a considerable advance over Layar, Wikitude and the other AR tools and apps. Google Visual Search, though experimental and incomplete, holds out the promise of being able to position the smartphone camera in front of a place or object and get information and data back by taking a picture or just by looking through the camera in the case of local businesses. 

This is a radical development for "mobile search" and is clearly useful in ways that most AR apps and tools are not. The only thing truly comparable in my mind, though on a more narrow level, is Amazon's "Amazon Remembers" (SnapTell) capability. Others may protest that Layar or other AR apps offer comparable or greater utility. But I don't agree.

Rather than attempting to be a complete substitute for the conventional search box or even voice search, the camera or "visual search" complements Google's other approaches. It works for objects, products and places that I'm right in front of:

  • A TV set
  • A business card (OCR)
  • A work of art in a museum
  • A business
  • A building or landmark

This is especially useful when I wouldn't know what query to type or it would be particularly cumbersome or awkward to enter a query, even for voice. One example might be a interesting building without an obvious address or other identifying markers . . . or a public sculpture where the artist's name or title weren't visible. And the business card/OCR capability (see video below) can't be accomplished at all via "search." 

Here's an explanatory video:

Picture 53

In the image above, the ability to quickly discover information about this Frida Kahlo painting through the camera is much more pleasurable and efficient than typing in the query "Diego Rivera and Frida Kahlo" and getting back a bunch of links (even images), then having to sift through them. 

Again this is the use case for AR: I'm right in front of the thing (or eventually the person) I want to know more about -- not 100 yards away. I'm realistically not going to scan the horizon for restaurants; there are more efficient ways to get information about where to eat on mobile devices. 

The hope of many of the AR competitors was to beat Google to the punch with the new experience. But now that Google has shown this functionality and the breadth of its ambition with "Visual Search" the hope of competitors has got to be data depth, quality or some social angle that Google would have difficulty doing. 

Google's long shadow online and now in mobile and its ability to bring together the search box with voice -- and now the camera -- makes it almost impossible to disrupt the relationship the company has established with the end users and search. Google is proving it recognizes the limitations of traditional search on mobile devices and can quickly adapt to the new medium. Indeed, Visual search is quite a bit more compelling than the Google PC search experience and even makes that latter appear quite primitive by comparison.

Perhaps only Microsoft is in a position to challenge Google on the multiple fronts that have now opened up in what we can expansively call "mobile search."