Why are some faces seen as truly expressionless, but others are inexplicably off-putting? What, exactly, makes us register a seemingly neutral expression as Resting Bitch Face? The researchers enlisted Noldus’s FaceReader, a sophisticated tool engineered to identify specific expressions based on a catalogue of more than 10,000 images of human faces. The software, which can examine faces through a live camera, a photograph or a video clip, maps 500 points on the human face, then analyzes the image and assigns an expression based on eight basic human emotions: happiness, sadness, anger, fear, surprise, disgust, contempt, and “neutral.”
To establish a baseline, Rogers and Macbeth first had FaceReader assess a series of genuinely expressionless faces. Those expressions registered about 97 percent neutrality, Macbeth said; the remaining three percent included “little blips of emotion” — a touch of sadness here, a hint of surprise there, but nothing significant.
The FaceReader software analyzes Kanye, one of the few men commonly associated with RBF. (Courtesy Noldus Information Technology)
Then they plugged in photos of RBF all-stars Kanye West, Kristen Stewart and Queen Elizabeth. Suddenly, the level of emotion detected by the software doubled to six percent.
One particular emotion was responsible for the jump: “The big change in percentage came from ‘contempt,’” Macbeth said.
