YouTube’s Facial Recognition a Possible Privacy Killer? [RUMOR]

A recent patent submission by Google has sparked rumors suggesting that facial recognition could be the next step for YouTube. The patent was submitted specifically for a technology that would go beyond facial recognition in photos and allow Google to recognize faces in video stream.

Here is the description of exactly how Google plans to accomplish this, as described in patent speak: “A method of identifying faces in a video includes the stages of: generating face tracks from input video streams; selecting key face images for each face track; clustering the face tracks to generate face clusters; creating face models from the face clusters; and correlating face models with a face model database.”

Got that? Just in case all that “clustering” does not quite make sense, here is the translation: The technology would essentially compile a database of your face using various angles captured from video streams that you are featured in. Google would then use that database to create a facial model that would allow them to recognize you and attach your name to videos floating around on a site like–oh, I don’t know–the #1 online video streaming site that happens to be Google-owned, one known as YouTube?

Just a year after Facebook introduced their own facial recognition technology to photos, and right on the heels of their acquisition of–which made their technology more accurate; Google seems to be developing a slightly more advanced version of a similar technology that would use moving images to essentially do the same thing.

This patent-pending technology brings to mind the ever-present concern of Internet privacy, which is raised whenever technology advances in a way that makes the almighty Internet able to track, locate or recognize its users without their knowledge or permission.

Believe it or not, there are still the few and the brave who seek to maintain some level of Internet privacy, and this feature could significantly hinder their efforts. We already know that Google is #1 in online stalking for advertising purposes, so putting this kind of technology in their hands is a serious cause for concern when it comes to online privacy.

Google’s technology would most likely emulate Facebook’s use of facial recognition in that although users can opt out of Facebook recommending their recognized image to friends for tagging purposes– users cannot prevent Facebook from using this technology to create a facial model based on the compilation of their photos.This is reminiscent of Google’s history of tracking users’ online movement without their knowledge through their Gmail accounts.

The difference, however, is that Facebook and Gmail users agree to terms and conditions and subsequently submit to all new developments that the sites undergo by simply creating an account. YouTube, on the other hand, is a whole other story due to the fact that unless you create an account you are not subject to the terms and conditions of the site or any technology they might use–you are simply an unattached third party viewer. The problem is that although you may not personally have an account, your friends who upload videos with you in them could be inadvertently allowing Google to start creating your facial model and adding your name to videos that match their compiled description–all without your knowledge or permission.

This could mean a serious invasion of privacy and future repercussions, especially considering that we all know that once something goes on the Internet it is forever searchable to, say, future employers. The concern is that YouTube’s use of this potential technology could result in Google compiling a complete video-centered online resume for you, without your knowledge.

So what do you think–are you worried about potential privacy invasion when it comes to facial recognition technology?


Comments are closed.