Slashdot Mirror
Google Begins Blurring Faces In Street View
mytrip notes a News.com article reporting that Google has begun blurring faces in its Street View service, which has spawned privacy concerns since its introduction last year. Google has been working for a couple of years to advance the state of the art of face recognition. Quoting News.com: 'The technology uses a computer algorithm to scour Google's image database for faces, then blurs them, said John Hanke, director of Google Earth and Google Maps, in an interview at the Where 2.0 conference...' Google wrote about the program in their Lat/Long blog."
This is the nice thing about living in a town no one cares about/knows about.
I cried real tears when Li Mu Bai died.
It's been awhile since a Google post on Slashdot has focused on the company improving our privacy. Good work!
"I zero-index my hamsters" - Willtor (147206)
jngpu gurz hfr fbzr erirefvoyr zrgubq yvxr ebg13, gubfr abbof
It would be cool if there were an option on sites like Facebook or Flickr to blur the faces on my photos for anyone but my friends.
With technology like this, I wonder how far away Google Image Search is from being able to search image content?
steampunk web design
Print a giant face over your storefront/building just to see what happens.
Damn...there goes my 15 minutes of fame.
My understanding is that people in public should have no expectations of privacy. Or is that just a U.S. thing? Furthermore, as their algorithms get better, will Google skip blurring the faces of famous people? They certainly have no expectations of privacy in public.
It must have been something you assimilated. . . .
You can't add pixels that aren't there, and an out of focus picture is effectively a lower resolution.
You can, however, apply statistical analysis and AI learning techniques to guess the likely locations of pixels. In that way, you can sharpen a photo somewhat, though it may be inexact. My understanding is that contextual analysis is the next step- if you have pictures of a person and a blurry person, and have more pictures of that person and less-blurry people, you can make predictions about who the fuzzy people are.
Of course, I wear a beard so that I'll always be fuzzy.
"The cup is in turn designed for holding hot or cold liquids, and has an open rim and closed base." --US Patent #5425497
They should have used Laughing Man logos. You blew it Google.
Could these enhanced algorithms be used to blur the faces of the hideous women I bring home from the bar? If not in real time, I'll accept them being blurred in my memory.
Google isn't blurring faces in the photos, but is actually blurring people's faces. Somehow, the Googlebotmobile blurs peoples' faces as it drives by, and so far no one has figured out a way to undo it.
Because, you know, the LAST thing I want to happen when I'm out on a public street is to be seen by millions of invisible people hiding in the Google van.
O HI, I FIXED UR POST, KTHX.
Looks like Google also cares a about horse privacy. That's really great! I woudn't want anyone recognizing my horse if he's caught doing something embarassing out in the street.
This article from a year ago shows that Google has had public implementations of facial recognition for some time. Simply appending &imgtype=face to a Google image search URL will just show images of faces.
Q: Option A or option B?
A: Yes.
"Wise men talk because they have something to say; fools, because they have to say something" - Plato
Why blur? Haven't we learned yet that the goal is no information, not less information? O.K., this is probably not one of those cases where someone will go to the trouble of trying to deconvolute the image. But really, just drop a white circle over the face and be done with it. Blurring gains nothing and leaves trace information.
"Uh... yeah, Brain, but where are we going to find rubber pants our size?" --Pinky
From a signal processing perspective, this is the same as convolving with a Gaussian. And if you take the Fourier transform of that blurred image, you get the transform of the image multiplied by the transform of the Gaussian (which is just another Gaussian). From there all you have to do is divide by this Gaussian, take the inverse transform, and walla, you have the desired non-blurred image. This is called a deconvolution, and I've written code to do this for an image processing class.
There are some caveats. You have to guess how blurred the image is - what focal length is and what not. Noise and compression can kill you, so you need to filter those out first (or limit your deconvolution filter to low frequency content). In addition at the edges of the image (or edge of the blur boundary) information is genuinely lost as the gaussian falls outside the boundary and is discarded.
Focus Magic is a commercial package that refocuses blurred images, and they have some interesting sample photos.
Oh, one other caveat, is that when you quantize the blurred image (assign each pixels a discrete, say 24-bit, value), you will also loose some information.
:) I was just surprised myself to learn that a blurred image is not the same as a lower resolution image, and so I thought I'd share.
Furthermore, I should mention that given the size of peoples faces, and the amount of blur that Google is likely to use, the entire blurred section will be near enough to the edge to loose significant information, so it is unlikely that much recovery will be possible.
So, nothing I said was really applicable to this situation
And the dial-up Internet access?
Don't waste CPU cycles blurring the image. Just past my face over everyone else's. I don't mind at all! Anyway, people who don't want to be recognised in public should know better that to leave home not wearing a burka.