Algorithm better than men at recognizing micro-expressions




Researchers at the Chinese Academy of Sciences have developed an algorithm that is as good as or better than men at recognizing micro-expressions. These are small, involuntary facial expressions that indicate emotions.

The training of the algorithm was challenging for the research team because the emotions have to be real and its expression by the observed to be suppressed. To achieve this, they left watching a group of subjects for a film that evokes certain emotions, while their faces were filmed with a high speed camera. They are instructed to not show their emotions. She was told that they do well, incentives, they must fill emotion shown by a large and boring form.

On the basis of the film provided the researchers a good idea of ​​what emotion at any time on the agenda. In this way the researchers could hear the algorithm learn what twitching in the face at the different feelings. As a starting point for the algorithm is a frame from the video containing no emotion in it.

The biggest challenge in building the algorithm lies in the size of the expressions. They are so small that the computer has difficulty recognizing them. As a solution, the researchers algorithm digitally enhance facial expression. Very slightly raised eyebrows be converted into an expression of extreme surprise. That makes it easier for the computer to identify the emotion. The image is also recorded in other color spaces to create emotions more visible. Besides rgb also be CIELuv and CIELab, and a new color space that the researchers tensor independent color space, or tics, call. This four-dimensional array makes the work of the algorithm as simple as possible.

micro-expressions algorithm
Image: MIT Technology Review

When the trained algorithm to compete with the human ability to identify micro-expressions, it is at least more or less equal on. When the algorithm and 15 human opponents are instructed to identify both micro-expressions in the above video recording as appoint, both sides scored about the same. When there are frames to be packed which predetermination which is shown a micro-expression, the algorithm outperforms the human opponents.

The researchers believe that the algorithm may be applied to such things as lie detection, mental health and law enforcement. The research was published in IEEE Transactions on Image Processing.


In: A Technology & Gadgets Asked By: [20441 Red Star Level]

Answer this Question

You must be Logged In to post an Answer.

Not a member yet? Sign Up Now »

Star Points Scale

Earn points for Asking and Answering Questions!

Grey Sta Levelr [1 - 25 Grey Star Level]
Green Star Level [26 - 50 Green Star Level]
Blue Star Level [51 - 500 Blue Star Level]
Orange Star Level [501 - 5000 Orange Star Level]
Red Star Level [5001 - 25000 Red Star Level]
Black Star Level [25001+ Black Star Level]