Your data. Your choice.

If you select «Essential cookies only», we’ll use cookies and similar technologies to collect information about your device and how you use our website. We need this information to allow you to log in securely and use basic functions such as the shopping cart.

By accepting all cookies, you’re allowing us to use this data to show you personalised offers, improve our website, and display targeted adverts on our website and on other websites or apps. Some data may also be shared with third parties and advertising partners as part of this process.

Background information

Electronic facial recognition will never be able to interpret emotions

Spektrum der Wissenschaft
27.4.2020
Translation: machine translated

Angry, happy, sad? One day, algorithms will be able to read our emotions from our faces. Researchers are now saying that this will probably never work.

The man in the picture - is he yawning? Is he singing? Is he perhaps even shouting? It's not that easy to read a person's facial expressions. Let alone draw conclusions about a person's state of mind. But this is exactly what computer algorithms for facial recognition should be able to do in the near future. Researchers led by computer scientist Aleix Martinez from Ohio State University have now explained that software will probably never be able to do this. The scientists reported at the annual conference of the American Association for the Advancement of Science that facial expressions would not reliably convey what a person is feeling. Because people can also disguise their facial expressions. In addition, certain facial expressions are valid for different emotions. Psychologist Lisa Feldman Barrett from Northeastern University, who also spoke on this topic at the science conference, cited studies showing that angry people actually only looked grumpy 30 per cent of the time. They would also make the same facial expressions when they were concentrating or reacting to a bad joke.

Another problem is caused by different cultural influences. One example: When someone opens their eyes wide and raises their eyebrows, this facial expression usually conveys surprise in Central European cultures. In some countries in south-east Europe and Anatolia, it is customary to simply say no with this expression.

The context counts

As Martinez and his colleagues go on to report, various companies have already developed algorithms that can deduce emotions from facial expressions. The researchers have tested these types of AI. The researchers' conclusion: "Some (companies) claim they can determine whether someone is guilty of a crime or not, whether a student is paying attention in class or whether a customer is happy with their purchase," says Martinez. "Our research has shown that these promises are complete nonsense." Such facial recognition programmes would prove to be particularly problematic if, for example, professional decisions are made on the basis of their results.

After all, it is the entire context that counts when reading a person's emotions, the researchers emphasise. So not only what facial expressions a person shows, but also their facial colour and posture as well as the place they are in.

And what about the man in the picture at the top? What is he really doing? He's obviously yawning.

Spectrum of science

We are partners of Spektrum der Wissenschaft and want to make well-founded information more accessible to you. Follow Spektrum der Wissenschaft if you like the articles.

[[small:]]

9 people like this article


User Avatar
User Avatar

Experts from science and research report on the latest findings in their fields – competent, authentic and comprehensible.


Background information

Interesting facts about products, behind-the-scenes looks at manufacturers and deep-dives on interesting people.

Show all

These articles might also interest you

  • Background information

    How critical can AI be? Interviewing artificial intelligence

    by Martin Jud

  • Background information

    Cute or creepy? This robot develops its character with your data

    by Stefanie Lechthaler

  • Background information

    Your smartphone knows where you’ve been – and that’s a problem

    by Oliver Herren

5 comments

Avatar
later