The annual Saudi Arabian camel beauty pageant was rocked with scandal recently, after some contestants had been found to have had botox injections to compliment their camel complexions.
With prize money well into in the millions, it’s hardly surprising that competition is smoking hot to be the most delectable dromedary of the day.
But there’s a serious side to this issue, not in camel competitions, but as to how a human face is read by machines.
Facial Recognition software is already widely used at passport gates the world over, but it’s also becoming a widely discussed topic in the world of law enforcement, recruitment and healthcare. Increasing reliance is being accorded to this technology; however, AI driven programs in this area might still struggle to give an accurate result if dealing with non-natural facial factors.
Many AI facial analysis algorithms gather data from changes in the movement of a person’s cheek, eyebrow and mouth muscles, along with potentially millions of other tics and tell-tale signs. As a result, if people’s faces have non-natural external accessories or treatment, such as spectacles or even Botox applications, results may be skewed from the norm.
Emotions like anger, sadness and happiness may be detectable, but could these be misread if other factors skew the results? The consequences could be far-reaching.
Read the controversial article here from the Guardian online,(Sept 2017), which reports that the AI proponent Michal Kosinsky claims to be able to accurately predict a person’s potential level of criminality and also assess their sexuality from facial analysis.
So it might not be too far in the future before job applicants or those being interviewed by police must remove their specs and declare any cosmetic surgery on a form.
The good news is that Fountech specialise in going the extra mile with AI, creating algorithms to take into account such additional factors into given parameters. It’s what we do, pushing AI one stage ever further, for better results, ahead of the game. Contact us to find out how we can help your organisation by integrating AI into your products and services.
April 2018 Published in Forbes