Stories
Slash Boxes
Comments

SoylentNews is people

SoylentNews is powered by your submissions, so send in your scoop. Only 18 submissions in the queue.
posted by martyb on Friday February 15 2019, @05:30AM   Printer-friendly
from the nowyouknowyouknow dept.

By now you're probably already aware of https://thispersondoesnotexist.com/

It is a website that uses Generative Adversarial Networks (GANs) to generate photos of people who do not exist. This story was sourced from https://www.inverse.com/article/53280-this-person-does-not-exist-gans-website

The article goes into some depth on how the researchers achieved this. But deepfakes of this nature present a problem. There are the obvious "safety concerns" for users of dating websites but these fakes are good enough to get on linkedin as well and perhaps fool an employer or other unaware person.

Fortunately, AI of any stripe, including GANs are not and never will be perfect and on this soylent exclusive, I wanted to take a moment to ring the alarm bells and explain how you can detect fakes like this.

AIs are not magical. They are merely complex heuristic and statistical systems, built and trained for a specific purpose. As such, they are only as good as their training data.

Obviously, the researchers had to source a training set from somewhere and it is clear that they used Facebook, Linkedin and Instagram.

As such, these photos, even the high quality ones all contain mistakes that AI would make (statistical mistakes), that a real human would not have and these mistakes are all tells.

Go ahead and go to the website, let it generate a photo. 8 out of 10 look perfect at first glance. I even found a few that looked like me and family members.
But remember people are describing these as "eerie"? Well there's a reason these look eerie, even the best ones.

That reason is from a statistical mistake in the eyes, specifically the pupils. Every one of these photos has at least two problems.

#1 is that the pupils are never dilated the same. In a normal healthy human the pupils dilate to the same extent, always. The only time they don't is when there is a brain injury such as acute head trauma or stroke. Since none of these people appear to be in a medical context where we might expect blown pupils, they appear creepy, deranged, crazy, brain injured etc.

#2 is that normal humans have no red in their irises. This appears to be a statistical mistake. Normal amateur photographers will frequently capture a concept called "red eye" where the flash of the camera reflects off the retina producing a red glow that we are all familiar with. The training set for this AI appears to have had a large number of red-eye photos in it and as a result there is red in the eye. But the red is not inside the pupils, it gets painted onto the irises. As a result, everyone appears to have toxic heavy metal poisoning (the usual cause of red splotches in the iris), in addition to traumatic brain injury.

There is also a third problem that is present in about 40% of the images. That mistake has to do with the hair. The way the hair is generated especially on men produces a "doll hair" effect, where hundreds maybe thousands of strands all pop up from clumps, in the same way it does on a doll. Unless the person has recently had hair transplant surgery, this is just not a thing that happens in real life.

So now you know. You'll be able to pick out the very best fakes this AI has to offer.


Original Submission

 
This discussion has been archived. No new comments can be posted.
Display Options Threshold/Breakthrough Mark All as Read Mark All as Unread
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
  • (Score: 2) by SemperOSS on Friday February 15 2019, @08:51AM (3 children)

    by SemperOSS (5072) on Friday February 15 2019, @08:51AM (#801464)

    I have yet to find eyes with red in the irises, and the pupils are, more or less, dilated the same. The hair is sometimes a bit off, but what I do find in many pictures are problems with symmetry.

    If you look at what should be near-symmetric features on each side of the face (i.e. not in the middle like nose or mouth), you'll see they are often less symmetrical than expected. The shape, slant and texture of the eyebrows often differ too much between the two sides; similarly with the ears, where it is shape and size. If both ears are visible and have earrings, they are rarely the same and often not very well defined. As much as people wear unmatched earrings in reality, it is not the norm still, I think.

    The problem I found most prominent, though, is that of glasses, when shown. They rarely show symmetry, and even a cursory glance says Frankenglasses on most of the pictures with glasses in them. Different shapes and sizes of lenses: the end pieces with the hinges are not in the same relative height to the lenses; they are not the same shape; and the temples (the arms going back over the ear) are not symmetrical and go at different angles behind the ear.

    These pictures — warts, flaws and all — are still amazing in many respects and show how far we have come since features like Final Fantasy in 2001.

    --
    I don't need a signature to draw attention to myself.
    Maybe I should add a sarcasm warning now and again?
    Starting Score:    1  point
    Karma-Bonus Modifier   +1  

    Total Score:   2  
  • (Score: 1) by SemperOSS on Friday February 15 2019, @09:08AM

    by SemperOSS (5072) on Friday February 15 2019, @09:08AM (#801468)

    I have been looking at some more pictures and realise that the thinner the frames of the glasses are, the bigger the problem, which obviously reflects the difficulty of the algorithm in separating the features of the glasses from the features of the face.

    And, not related to symmetry, I found a picture of a boy with huge faults in it. Half of his right ear was separated from the head, an off-coloured splotch on his left cheek, a red line like a thread in his right cheek and a big off-coloured blob going from his right cheek down to his chin, to name the most prominent. This was well beyond uncanny valley and far into oops territory.

    --
    I don't need a signature to draw attention to myself.
    Maybe I should add a sarcasm warning now and again?
  • (Score: 3, Interesting) by looorg on Friday February 15 2019, @10:53AM (1 child)

    by looorg (578) on Friday February 15 2019, @10:53AM (#801485)

    I pulled up 10 images, perhaps not enough of a sample size but still better then just looking at one image, and didn't notice any of the red-eye irises either. The hair looked fine on them to, more or less. I think the only reason I think they look fake is that I have been told that they are fake. The issues otherwise seems to be more if I look at each picture for a really long time they do start to seem a bit off but I can really tell if that is just in my mind. Sure I can say that the pupils are different sizes and that the facial expressions sometimes looks weird and as noted the symmetry sometimes seems a tad off. But in general or at a glace for the most of them I can't tell the differences between fake person and real person. If anything they should add in real images to and have an option for people to "vote" fake or real just to see if people can actually tell the difference.

    • (Score: 2) by Runaway1956 on Saturday February 16 2019, @02:08PM

      by Runaway1956 (2926) Subscriber Badge on Saturday February 16 2019, @02:08PM (#802032) Journal

      Symmetry. I'm looking at one right now, of a female. She has a fat woman's jowl on her left side, a slightly mismatched cheek without a jowl on her right. Odd . . .