News for hackers. Stuff that matters.
Re:Facebook's '10 Year Challenge' Is Just a Harmless Meme–Ri... by flocial on Sunday January 20, 2019 @11:04AM (#18953600)
Facebook already has access to a larger repository of photos going back over a decade and all the years inbetween along with decent face recognition to create a much bigger dataset than resorting to a hashtag challenge. But I guess that wouldn't be newsworthy.
Re:Facebook's '10 Year Challenge' Is Just a Harmless Meme–Ri... by reaperducer on Sunday January 20, 2019 @12:13PM (#18954087)
Facebook already has access to a larger repository of photos going back over a decade and all the years inbetween along with decent face recognition to create a much bigger dataset than resorting to a hashtag challenge. But I guess that wouldn't be newsworthy.

This was addressed in the article:

In various versions of the meme, people were instructed to post their first profile picture alongside their current profile picture, or a picture from 10 years ago alongside their current profile picture. So, yes: These profile pictures exist, they’ve got upload time stamps, many people have a lot of them, and for the most part they’re publicly accessible.

But let's play out this idea.

Imagine that you wanted to train a facial recognition algorithm on age-related characteristics and, more specifically, on age progression (e.g., how people are likely to look as they get older). Ideally, you'd want a broad and rigorous dataset with lots of people's pictures. It would help if you knew they were taken a fixed number of years apart—say, 10 years.

...

In other words, it would help if you had a clean, simple, helpfully labeled set of then-and-now photos.

Re:Facebook's '10 Year Challenge' Is Just a Harmless Meme–Ri... by Kylekramer on Sunday January 20, 2019 @12:59PM (#18954399)
It tries and fails to address it. I could barely count the number of zeros I would have to put in front of the 1 in the percentage of Facebook's photo data this meme covers. And a good portion of their dataset will have EXIF timestamps. Training an algorithm on the meme only would be insane waste of their data set.
Re:Facebook's '10 Year Challenge' Is Just a Harmless Meme–Ri... by bbrian on Sunday January 20, 2019 @02:05PM (#18954875)
> their dataset will have EXIF timestamps

It used to be the case (still is?) that the dates on uploaded photos weren't applied to the photo album. I remember having to go through holiday snaps and change the date on each from the upload date to the actual date. The images were also resized down from what was uploaded.

So, if they have the originals with the full EXIF data, I'd like to be able to use that for my old photos!

Re:Facebook's '10 Year Challenge' Is Just a Harmless Meme–Ri... by bigbugbag on Sunday January 20, 2019 @02:56PM (#18955212)
Chances are they have and you won't be able to use them.

Why would it be so ? Because it profits facebook which is the only reason facebook exists. It profits them to have original with EXIF for data mining and you gave them permission to do so while also giving them the data, and it profits them to not make them available to you to save on bandwidth and processing costs.

Re:Facebook's '10 Year Challenge' Is Just a Harmless Meme–Ri... by Setepenre on Monday January 21, 2019 @07:44AM (#18959968)
eh, I think not. There is a reason why they resize the photos. To save space. Even at facebook scale the amount of space they save by doing this must be enormous.

Additionally, images used in AI are usually scaled down a lot more, 224x224 for something like resnet50. which means that they do not need your high quality original and the smaller one they generated are fine.

Re:Facebook's '10 Year Challenge' Is Just a Harmless Meme–Ri... by cmmartin on Sunday January 20, 2019 @02:00PM (#18954830)
They don't need everyone to train their AI to detect aging. They don't even need a substantial sample size. All humans age in relatively similar ways.
Re:Facebook's '10 Year Challenge' Is Just a Harmless Meme–Ri... by baroffoos on Sunday January 20, 2019 @08:16PM (#18956785)
Why would they even bother with this when they already have much better data though.
Re:Facebook's '10 Year Challenge' Is Just a Harmless Meme–Ri... by DoctorOetker on Sunday January 20, 2019 @01:57PM (#18954809)
perhaps they forgot to split a test set :)

while I'm sure they already have a great dataset going back far, most of that dataset will be a small fragment of the world that already had internet 10 years ago, in the mean time a lot of people in other parts of the world have come online and may have old pictures of themselves

Re:Facebook's '10 Year Challenge' Is Just a Harmless Meme–Ri... by jonlucc on Sunday January 20, 2019 @12:59PM (#18954402)
Lots of people have said this, but Facebook has upload directly from phones to their server, right? So I would be shocked if they aren't trusting the time tags from direct uploads and metadata set by the camera (which they strip before displaying, but surely they ingest it) more than they trust upload times.

Also, people tag themselves all the time, and if you upload a photo with people, you can tell quickly that they already identify people with high accuracy, because they suggest tags for you.

Re:Facebook's '10 Year Challenge' Is Just a Harmless Meme–Ri... by bigbugbag on Sunday January 20, 2019 @02:59PM (#18955239)
Seems you're missing one thing here, rewind 10 years back and the iphone 3G got released at the end of that year so not many people were using phones to take pictures and even less uploading them as their data plan did not allow when coverage allowed to do so.
Re:Facebook's '10 Year Challenge' Is Just a Harmless Meme–Ri... by nullymcnull on Sunday January 20, 2019 @03:47PM (#18955471)
You're really reaching now. People were regularly using phone cameras long before smartphones. By the time smartphones were a few gens in it was ultra common, and a big benefit of them (that even non-technical people could understand) was that you could use wifi whenever possible and avoid data charges (which really weren't much higher then than now in most markets; "unlimited" plans were more common too). Also the 3g has a 2mp camera, so we're talking about pictures that are 2-3MB at most. The suggestion that people were shy about uploading relatively small photos to Facebook circa ~2008, and that this supports this flimsy story in any way, is sheer nonsense. I'm no fan of Facebook - am a long-term outright refusenik actually - but the conspiracy theories are getting out of hand. There is zero substance to this article, it's wildly speculative clickbait.
Re:Facebook's '10 Year Challenge' Is Just a Harmless Meme–Ri... by ricardobeat on Sunday January 20, 2019 @03:58PM (#18955547)
In 2008 Facebook itself was in its infancy. Orkut was still the largest social network, only toppled around 2011. The majority of the world was definitely not uploading any phone pictures anywhere.
Re:Facebook's '10 Year Challenge' Is Just a Harmless Meme–Ri... by nullymcnull on Sunday January 20, 2019 @05:14PM (#18955962)
> In 2008 Facebook itself was in infancy

That's quite an overstatement, to put it very mildly. Facebook was allowing open non-.edu signups by 2006, and the buzz around it from it's school success was immense. By 2008 it certainly wasn't seeing a critical mass of boomers and other late(st) adopters, but it was still huge by any measure - 100 million users, and growing with unprecedented speed.

People were absolutely already uploading phone pictures to FB and other sites by then; I think there may even have been Facebook apps shipping on non-smartphones by that time, it was one of the earliest things carriers used to flog data plans. I agree that "the majority of the world" wasn't uploading phone pictures anywhere by then, but then I'd be surprised if that rather high bar has been reached today either.

Re:Facebook's '10 Year Challenge' Is Just a Harmless Meme–Ri... by ricardobeat on Monday January 21, 2019 @02:14AM (#18958343)
The matter at hand is collecting user pictures for mass scale machine learning. I and everyone I know didn’t even join FB until 2011-12. 100m is nothing compared to the current 2.2B user base who is posting annotated 10 year old pictures of themselves. This is to counter the parent comment that “they already have this data anyway”, not an absolute statement on FB growth.
Re:Facebook's '10 Year Challenge' Is Just a Harmless Meme–Ri... by matchagaucho on Sunday January 20, 2019 @12:55PM (#18954375)
It seems, at least, Microsoft is way beyond needing a 10-year challenge training dataset.

Their "how old" facial analysis app has been around a few years and is remarkably accurate.

https://www.how-old.net/

Re:Facebook's '10 Year Challenge' Is Just a Harmless Meme–Ri... by zmk_ on Sunday January 20, 2019 @01:50PM (#18954767)
It is actually not. I've been labelled anything between 22 and 65. It is not only wrong, but also not accurate about it (large variance).
Re:Facebook's '10 Year Challenge' Is Just a Harmless Meme–Ri... by macintux on Sunday January 20, 2019 @12:08PM (#18954059)
As has been pointed out, people identifying themselves makes it easier to figure out which 10-year-old and current photos are actually:

* themselves

* 10 years old

Re:Facebook's '10 Year Challenge' Is Just a Harmless Meme–Ri... by justapassenger on Sunday January 20, 2019 @12:44PM (#18954307)
And of course no one would lie about it!
Re:Facebook's '10 Year Challenge' Is Just a Harmless Meme–Ri... by macintux on Sunday January 20, 2019 @12:59PM (#18954406)
The percentage of people lying about it, at least before the stage where people start piling on with jokes, is probably much lower than the general noise in the data set if you were to try to find the photos you want without help.
Re:Facebook's '10 Year Challenge' Is Just a Harmless Meme–Ri... by Lerc on Sunday January 20, 2019 @02:59PM (#18955238)
That too is data. Those outliers just get sent to the not-to-be-trusted watch list.
Re:Facebook's '10 Year Challenge' Is Just a Harmless Meme–Ri... by universenz on Sunday January 20, 2019 @12:48PM (#18954323)
Jokes on you, Facebook is actually training their AI for deep fakes!
Re:Facebook's '10 Year Challenge' Is Just a Harmless Meme–Ri... by corobo on Sunday January 20, 2019 @11:12AM (#18953645)
That is assuming it's Facebook training the algorithm
Re:Facebook's '10 Year Challenge' Is Just a Harmless Meme–Ri... by acct1771 on Monday January 21, 2019 @12:39AM (#18957915)
Just say what you're saying. It's Feds. And they're the biggest social influencers around.
Re:Facebook's '10 Year Challenge' Is Just a Harmless Meme–Ri... by reaperducer on Sunday January 20, 2019 @12:15PM (#18954098)
Facebook denies involvement. But Facebook is no longer worthy of blind trust.
Re:Facebook's '10 Year Challenge' Is Just a Harmless Meme–Ri... by BurningFrog on Sunday January 20, 2019 @01:11PM (#18954490)
This is a circular argument.
Re:Facebook's '10 Year Challenge' Is Just a Harmless Meme–Ri... by 0xfffff on Sunday January 20, 2019 @02:51PM (#18955176)
did you read the article? thats even addressed
Re:Facebook's '10 Year Challenge' Is Just a Harmless Meme–Ri... by JaleDarvis on Sunday January 20, 2019 @11:08AM (#18953624)
Access to what?
Re:Facebook's '10 Year Challenge' Is Just a Harmless Meme–Ri... by berbec on Sunday January 20, 2019 @11:09AM (#18953630)
a larger repository of photos going back over a decade