MIT Computer System Spots The Difference Between Real And Fake Smiles
Researchers find that most people smile when they're frustrated but it is difficult for human observers to differentiate these from smiles of joy.
Researchers at MIT have found that most people smile when they’re frustrated, and have programmed computers to differentiate between these smiles and smiles of delight better than human observers can. Graduate students Ehsan Hoque and Daniel McDuff, and professor Rosalind Picard, conducted experiments at MIT’s Media Lab, which found that 90% of people smiled when presented with a frustrating task.
Still images showed little difference between these frustrated smiles and the delighted smiles produced when viewing a video of a cute baby. However, video analysis showed that the progression of these two kinds of smiles was quite different, with the happy smiles building up gradually, and the frustrated smiles appearing quickly but fading fast.