Neural Net Trained on Mugshots Predicts Criminals

Alternately: criminal prosecution is targeted at people who "look like" criminals; thus, the NN is just selecting those people we think look like criminals, rather than any inherent criminality.

Their method had a 89.5% success rate, which might seem great, but is pretty much worthless in real life. The US has the highest incarceration rate, so we can use the US incarceration rate as an upper bound for the probability of randomly selecting a criminal from the population (716 per 100k, P=0.00716). This means that if we apply the same method at random to members of the general population, there&#x27;s actually at most a 5.79% chance that the result of &quot;criminal&quot; is accurate.<p><pre><code> Maths: Let Pc = probability of criminal = 0.00716 Let Pt = probability of test being accurate = .895 Probability of criminal given criminal conclusion = Pc * Pt &#x2F; (Pc * Pt + (1 - Pc) * (1 - Pt)) = 0.0579</code></pre>

Give this to a despot to train on his political enemies (or ethnic&#x2F;religious minorities), and you suddenly have a very good NN for condemning innocent people to jail.<p>Research should continue into this, but it&#x27;s worth remembering that the &quot;criminals&quot; being trained on aren&#x27;t necessarily <i>bad people</i> in the moral sense. They&#x27;re merely the recipients of judgment by some third entity (in this case, China&#x27;s legal system).

&quot;Their method is straightforward. They take ID photos of 1856 Chinese men between the ages of 18 and 55 with no facial hair. Half of these men were criminals.<p>They then used 90 percent of these images to train a convolutional neural network to recognize the difference and then tested the neural net on the remaining 10 percent of the images.<p>The results are unsettling. Xiaolin and Xi found that the neural network could correctly identify criminals and noncriminals with an accuracy of 89.5 percent.&quot;

This doesn&#x27;t &quot;predict criminals&quot;, it predicts those who will be convicted of a crime, which is not the same thing. Suppose that people have an unconscious prejudice against those with eyes set close together. They will be disproportionately convicted, and this neural net will find the correlation.

&#x27;Criminals&#x27; being defined as people who have been convicted by a judicial system that is full of bias.

This concept is troubling, uncomfortable, and could potentially be the basis for some very bad policy but none of that is a reason to dismiss it outright. If these correlations are real, it&#x27;s worthwhile to find out more about it with an open mind.

I bet you could train a neural net to detect the exact moment when the AI bubble has jumped the shark.

Well what comes around, goes around or such. I remember during my studies (Literature, Culture and such) to having read of methodologies used in the 18th century to detect criminals by their physiological features.<p>On person doing these studies was for example Francis Galton (who by the way did quite a mix of things from eugenics to statistics and &quot;the wisdom of the crowd&quot;)[1].<p>Lots of things, coming from the old times into the modern times like Physignomy [2] that was also used for racial identification&#x2F;discrimination in the 20th century.<p>Have fun walking deeper into that rabbit hole of history. What I take from that is, that bad ideas never die, even if science was able to debunk them.<p>Or as @thechao already said:<p>&gt; Alternately: criminal prosecution is targeted at people who &quot;look like&quot; criminals; thus, the NN is just selecting those people we think look like criminals, rather than any inherent criminality.<p>[1] <a href="https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Francis_Galton" rel="nofollow">https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Francis_Galton</a> [2] <a href="https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Physiognomy" rel="nofollow">https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Physiognomy</a><p>[Edit] Formatting

The paper&#x27;s conclusion claims that &quot; Furthermore, we have discovered that a law of normality for faces of non- criminals. After controlled for race, gender and age, the general law-biding public have facial appearances that vary in a significantly lesser degree than criminals.&quot;<p>Given the above I move that the title of this post is changed to &quot;Neural Net trained on mugshots confirms the findings of Phrenology&quot;.

So were faces of e.g. dodgy GFC bankers included in that pool? Or are we just talking petty criminals here.

I looked at this paper the other day and it looked to me like the non-criminal faces examples were men with shirt collars, whereas the criminal examples were men in t-shirts. If they got above random accuracy, then I wonder if they simply overfitted on that, and the lighting and colour differences between the two styles of photos.

Did not think this would be the year phrenology made a comeback. . .

I wonder if it is the same before and after one becomes a criminal. Being a criminal does not mean they don&#x27;t regret or feel guilt. Perhaps that is what is detected.

&quot;In other words, the faces of general law-biding public have a greater degree of resemblance compared with the faces of criminals, or criminals have a higher degree of dissimilarity in facial appearance than normal people&quot;<p>Hmm... so.<p>You look weird -&gt; people treat you worse -&gt; you don&#x27;t feel like working with them -&gt; higher chance of being a criminal.<p>I know that&#x27;s a giant leap in reasoning, but that was the first thing that came to mind.

Could it be used to identify those likely to commit election or securities frauds? Or maybe those likely to poison an entire city by ruining their water supply?

Previous discussion on arXiv paper:<p><a href="https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=12983827" rel="nofollow">https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=12983827</a>

&quot;They take ID photos of 1856 Chinese men [...]. Half of these men were criminals.&quot;<p>Because half of the general male population is criminal, of course.<p>The accuracy rate would be very different with a training&#x2F;testing sample that takes the base rate of criminality into account.

Let me guess... the NN was trained to spot tattoos? I bet you tattoos and crime correlate.

Phrenology is still phrenology even if its digital.

The future of this would make a great film...