Personality recognition (just by looking at you)?

[rslade]

Creepy social engineering is one thing.  Basing law enforcement, physical security, investigations, and job interviews on highly questionable premises is quite another.

Enter Faception.

Faception claims to be able to "reveal personality from facial images" and "dramatically improve public safety, communications, decision-making, and experiences."  How?  Well, after some buzzword filled marketing jargon about "first-to-technology and first-to-market with proprietary computer vision and machine learning technology" and mention of the magic word "biometrics," if you persist you may be able to find the theory behind the technology.  It seems to boil down to the following logic:

1) DNA can determine (certain) personality traits (sometimes to a significant extent).  (This is true, with the provisos I've put in parentheses.)

2) DNA can determine how you look.

THEREFORE:

Your personality is determined by how you look.

(Finding the flaws in this argument is left as an exercise for students of logic.)

I am inescapably reminded of the "bomb detectors" sold to Afghani and Iraqi security forces that had no detection capabilities at all, and caused large numbers of deaths.  That's on the false negative side.  The potential damage caused on the false positive side are likely considerably greater ...

[CISOScott]

So the old Psychology debate nature vs nurture but in the next-gen, artificial intelligence type of world. We have now moved age-old discussions to the cloud!

So does the article say you are locked into how you look or are you able to morph into something else by the way you act or perceive yourself? I wonder how much of it is causational though? If you have a sad life therefore you frown more, so the frown muscles get stronger than the smile muscles and it changes your facial appearance. Is that what they are picking up on? Did they take into account different cultures and races?

There seems to be some scientific studies to back it up as I remember in my psychology classes there being a study on people with multiple personalities that involved different facial expressions or mannerisms when the person changed to a different person in their mind. I have also seen cases where, in the case of serial killers or serial rapists where the demeanor of the attacker would change and they would become more stoic and develop a "far-away" look right before they were about to commit their crimes or attacks. Kind of like the mind was shutting down the human warmth, and caring part of the individual and getting ready to turn on the violent, uncaring part of the individual so they could perform their attack.

I would be wary to say that detection would be 100% accurate because people change and life can through circumstances at you that you may not have planned for. It would be interesting to see what their controlled study groups looked like that they gathered their algorithmic data from.

[rslade]

> CISOScott (Contributor II) posted a new reply in Industry News on 07-16-2018

> I have
> also seen cases where, in the case of serial killers or serial rapists where the
> demeanor of the attacker would change and they would become more stoic and
> develop a "far-away" look right before they were about to commit their crimes or
> attacks.

I wouldn't want to have to be the one responsible for collecting the image data for
doing machine learning on that kind of thing ...

====================== (quote inserted randomly by Pegasus Mailer)
rslade@vcn.bc.ca slade@victoria.tc.ca rslade@computercrime.org
Literature is the art of writing something that will be read
twice; journalism what will be read once. - Cyril Connolly
victoria.tc.ca/techrev/rms.htm http://www.infosecbc.org/links
http://blogs.securiteam.com/index.php/archives/author/p1/
http://twitter.com/rslade

[DAlexander]

Plastic surgeons are about to get rich(er)!!

[Baechle]

---

@DAlexander wrote:

Plastic surgeons are about to get rich(er)!!

---

<sarcasm>I can't wait until they refine it so that it can detect even the most minute deviations in the skull in order to determine how risky someone is of being fraudulent or committing a crime. </sarcasm>  

History lesson on Phrenology: Neuroscientists put the dubious theory of ‘phrenology’ through rigorous testing for the first time

[CISOScott]

---

@rslade wrote:

I wouldn't want to have to be the one responsible for collecting the image data for
doing machine learning on that kind of thing ...

---

Yes that is why there is a high dropout/burnout rate in psychology and certain law-enforcement positions. You get to see past the dirty underbelly of society and it is not a pretty sight. When you see what some human beings do to other human beings it is really heartbreaking. it is like bullying on steroids times 1000. There actually is a disconnect of human emotions which I am sure can be picked up upon by a machine that is measuring everything.

[Baechle]

Ken,

This is undergraduate level Psychology/Sociology/Criminology 101.

---

@CISOScott wrote:

---

Yes that is why there is a high dropout/burnout rate in psychology and certain law-enforcement positions. 

There are different schools of thought here in what causes a person to behave a certain way.  In the Italian School of Criminology, Cesare Lombroso proposed the concept of biological positivist criminology.  This was the idea that genetics predispose a person to social deviance.  In this iteration the theory was proposed as the sole reason a person turned to deviant behavior (as opposed to a combination of factors).

There was a recent study by Dr. Barnes of UT Dallas along with Dr. Beaver from Florida State University, and Dr. Boutwell of Sam Houston State University that conducted a Twin Method study against life long offenders, adolescent offenders and abstainers.  I have to admit that the data is extraordinarily confusing, and when I go over it a few times, seems to contradict itself between the different groups.

What I think this study actually shows is that there is a biological tendency to be vulnerable to different types of physical, emotional, or mental stress.  For example, are some people more predisposed to be influenced by peers or need a "community" even if that community is a deviant sub-culture?  Are some people more biologically resilient to physical stress vs mental stress? 

I'll have to reread the study again, but I also think that the study failed to sufficiently exclude twins that were raised in similar environments as contributing to deviance.  What's more is that the research wasn't able to identify what genetic marker existed that made someone more vulnerable to a deviant set of life choices, only the correlation between maternal twins where neither, one, or both chose a deviant lifestyle.  And even more distressing is that at least 30% of the time genetics didn't correlate at all for those that chose a deviant life style.

Are we ok with a 30% false positive rate?

Sincerely,

EB