Computer Detects Fake Human Expressions Better than People

Faked grimaces of pain and other false winces may fool fellow humans, but not a computer, according to a new study.

Writing in the journal Current Biology, researchers from University of California San Diego and the University of Toronto report that a computer system can spot real or faked expressions of pain more accurately than real people can.

"The computer system managed to detect distinctive dynamic features of facial expressions that people missed," said lead study author Marian Bartlett, a research professor at UC San Diego's Institute for Neural Computation. "Human observers just aren't very good at telling real from faked expressions of pain."

When a human feigns a facial expression, it may be enough to fool another human, but the the computer system's pattern-recognition abilities are better at detecting whether those expressions are real or fake, said senor study author Kang Lee, a professor of child studies at University of Toronto.

Humans, it turns out, are not very good at distinguishing false grimaces of pain from genuine ones. The researchers found that humans preformed no better than random chance when discriminating between expressions of pain, and even after training could only improve their accuracy to 55 percent.

The computer system, meanwhile, could do so with 85 percent accuracy.

"In highly social species such as humans faces have evolved to convey rich information, including expressions of emotion and pain," Lee said in a statement. "And, because of the way our brains are built, people can simulate emotions they're not actually experiencing - so successfully that they fool other people. The computer is much better at spotting the subtle differences between involuntary and voluntary facial movements."

The giveaway from from the mouth, and how and when it opens. Fakers' mouths open too regularly and with less variation, the researchers said, noting that subsequent studies will look into whether over-regularity is a general feature of fake expressions.

Bartlett noted that the computer expression detection system could be used in a variety of real-world scenarios including homeland security, psychopathology, job screening, medicine, and law.

"As with causes of pain, these scenarios also generate strong emotions, along with attempts to minimize, mask, and fake such emotions, which may involve 'dual control' of the face," Bartlett said. "In addition, our computer-vision system can be applied to detect states in which the human face may provide important clues as to health, physiology, emotion, or thought, such as drivers' expressions of sleepiness, students' expressions of attention and comprehension of lectures, or responses to treatment of affective disorders."