Facial recognition coming to Google Glass, sort of
- 19 December, 2013 06:12
For months, Google has steadfastly maintained that it will not allow facial-recognition apps onto Google Glass. No problem. A third-party developer, Stephen Balaban, will merely offer his own FaceRec app for sideloading onto Glass devices.
Nevertheless, there are several limitations to that approach: first, the Glass itself must be "rooted," which will slice the limited number of Glass devices down into an even smaller number. And even if this is done, the user will be forced to create his own image database, or write a custom script to tap into Facebook's image library--which violates Facebook's terms of service.
Balaban has also developed a so-called "Lambda Hat" to overcome another technical limitation with Glass and facial recognition. While operating, Glass only has the battery power to run FaceRec for a couple of hours. The "hat" is a standalone device with a longer battery life.
Nevertheless, Balaban and his startup, Lambda Labs, tells Forbes that the app will be released at the Chaos Communications Conference in Berlin, which begins on December 27. (The conference, dedicated to hacking, has included lockpicking competitions, and this year will dedicate a track to "food hacking.")
When it comes to the privacy elements of Google Glass, people usually view it as a perpetually-operating spycam, or else as a valuable tool to identify people and places and provide them with additional context (as say, most CRM software attempts to do with electronic communications). Balaban obviously falls into the latter camp.
So far, the app appears just to snap photos every 10 seconds or so, correlating them with a particular location. As such, FaceRec appears just a bit more than the sort of "lifestreaming" concepts that bubbled up a few years ago, measuring when and where users moved, and possibly who they spoke to. (Microsoft's MyLifeBits is another example.)
Glass, of course, has the potential to go far beyond that, superimposing all sorts of information on top of the subject's image. In June, that sort of scenario prompted more than 30 privacy commissioners to send Google a letter outlining their concerns, and even prompted a Seattle bar to ban patrons from wearing them. (Google Glass users say they can be trusted with others' privacy.)
From a productivity standpoint, there's a strong case to be made that a facial-recognition app authored by Google--or Microsoft, or Facebook--could be incredibly useful...but only with the proper privacy restrictions. Facebook, to its credit, has reversed course from its laissez-faire attitude toward privacy to at least provide controls to let users restrict posts and opt out of being identified, or tagged, in other's photos. If a Facebook or Google-designed system were put in place with the ability to opt out of being recognized, by default, it's possible that facial recognition would receive a warmer welcome. Controls to allow the user to toggle recognition on and off, such as while attending a conference, would be even better.
The problem with facial recognition is that it's not a black-and-white issue. Placing control of the technology within a global multinational corporation whose business is built on sharing and selling your data scares many, and for good reason. But putting that technology into the hands of a few hackers may also be cause for worry. And then there's the case of the technology itself: being identified by a long-lost friend is one thing. Being recognized as, say, a doctor who performs abortions is quite another.
For now, it sounds like the facial recognition being talked about by Balaban doesn't sound too worrying, even for privacy wonks. Balaban's challenge, if he accepts it, will be to build out the technology in a way that's both useful and, hopefully, respectful of the issues.