By Bekki Bushnell, Head of Business Development
Beware those of you with a slightly curvier upper lip – you could be mistaken for a criminal.
A paper has been released asking if artificial intelligence can detect if a human can be a criminal by analysing their facial features.
So if you have a smaller mouth and closer-set eyes, don’t be surprised if the waitress at Starbucks gives you a slightly more cautious glance before handing over your morning coffee.
Researchers from Shanghai Jiao Tong University released the study on arXiv, an online print journal that is handily open-source, but has not yet been officially published.
The aforementioned three key features, the researchers Xiaolin Wu and Xi Zhang believe, are what set criminals apart.
Their claims are slightly fanciful: “Unlike a human examiner/judge, a computer vision algorithm or classifier has absolutely no subjective baggages, having no emotions, no biases whatsoever due to past experience, race, religion, political doctrine, gender, age, etc, no mental fatigue, no preconditioning of a bad sleep or meal.”
Just make sure you’ve had Botox on the day of conviction then?
The study involved more than 1,800 faces of Chinese men aged 18-55 which were ‘controlled’ to account for face, gender, age and facial expressions. More than 700 of these images fed into the programme were of criminals (it should be said were not mugshots). And what was thrown out? This kind of facial bias.
According to a summary of the piece on the Daily Mail, a user on Hacker News said this should be “taught in class as a counterexample” of how to use AI.
Just this week we have seen Prime Minister Theresa May discuss artificial intelligence and the workforce of tomorrow. Though some would discount her opinion for terming one social media giant “the Facebook”.
There is no doubt that AI will continue to make the national news agenda. The discussions around how it will impact on Britain’s workforce has really come to the fore and it is a fascinating debate.
But maybe Britain’s judges can sleep a little easier knowing their jobs aren’t under threat…for now…