Scientists have been trying to devise an accurate way to detect that someone is lying or cheating on a target for decades, and now, thanks to artificial intelligence, scientists think they may be close to meeting the target by a significant percentage.
Artificial intelligence can detect deception
According to what was published by the British “Daily Mail”, citing the leading journal Brain and Behavior, a team of scientists succeeded in creating sensors that succeeded in reading the faces of a number of volunteers, and monitored and monitored slight changes in facial movement while they were telling lies or truthful facts.
The AI system was able to tell if someone was lying with an accuracy of 73%, which is slightly lower than the polygraph test, which reached an accuracy degree of 80%, but scientists say that this percentage was achieved at a very early stage of the experiment, which means It will be improved in the future.
Scientists from Tel Aviv University expect that in the future AI-equipped cameras could be used at airports, in online job interviews or in police interrogations of suspects to see if someone is lying.
Devices are superior to humans
Previous research revealed that humans can tell a lie from the truth with an estimated rate of about 55%, while a polygraph is accurate up to 80%, but these percentages are not enough to accept the results of a lie detector test as evidence in a court of law, so researchers around the world are working on new solutions.
The researchers used machine learning and artificial intelligence programs to quickly analyze tiny changes in muscle movements while lying, including small movements of the cheek muscles and eyebrows.
Measurement of facial muscle torsion.
Facial movements were measured using stickers printed on soft surfaces containing electrodes capable of monitoring and measuring nerves and muscles. The researchers attached the stickers to the muscles of the cheeks near the lips and the muscles above the eyebrows.
The study is based on the assumption that facial muscles twist when a person lies, and that until now no electrodes are sensitive enough to measure these torsion, so the researchers trained a machine-learning program to identify lies based on EMG signals from the electrodes. By applying this method, an accuracy of up to 73% was achieved, and it may not be ideal, but it is much better than other techniques used for the same purpose.
The researchers expect that the innovative AI program will have “dramatic effects in many areas of life”, as the electrodes could be available in abundance in the future, and then could be used alongside video software to detect lies by watching facial muscle movements.
Thus it will be possible to use the innovative technology in the areas of “banks, interrogations in police departments, at airports or in online job interviews, [إذ] High-resolution cameras trained to identify facial muscle movements will be able to [تحديد ما إذا كانت] Truthful statements from lies.”
Criticisms in the past
The idea of using artificial intelligence to detect liars has been criticized in the past. For example, in 2018, researchers at Manchester Metropolitan University suggested that AI could be used to detect whether people are lying to border security by analyzing their precise gestures, but Professor Ray Paul, Professor of Criminal Investigation at the University of Derby, criticized the idea, saying it was a project. “It lacks credibility, because there is no evidence that observing small changes on people’s faces is an accurate way to measure lying.
Professor Ball said: “They [علماء جامعة مانشستر متروبوليتان] They delude themselves into thinking they will ever be highly effective. They waste a lot of money,” because that technology is based on a fundamental misunderstanding of what human behavior can be like when they are honest and deceptive.