Algorithm engineered that can spot hidden feelings

Share this on social media:

Machines that can tell how a person is feeling by recognising almost imperceptible facial expressions could soon be a reality thanks to advances in machine vision algorithms. Scientists at the University of Oulu in Finland have developed an algorithm capable of spotting microexpressions, a form of facial expression that only occurs fleetingly and reveals true emotions even when the person is trying to hide them.

The work, led by Xiaobai Li at the University of Oulu, recorded the expressions of 20 individuals on a camera at 100 frames per second. The group watched a series of videos designed to invoke strong emotions but were also given an incentive to avoid displaying any emotion. Microexpressions tend to occur when the individual is trying to hide what they're really feeling.

The camera picked up 164 microexpressions from the group by recognising when the expression changed from frame to frame. These were then linked to the emotions exhibited in the videos and the images used to train the algorithm.

To overcome the inherently small degree of facial movement exhibited in microexpressions, the scientists algorithm magnified the parts of the face that moved before classifying the emotion.

The team’s algorithm significantly outperformed humans in recognising microexpressions in a comparative test. According to the team, the system is the first one that’s been tested on a spontaneous microexpression data set.

Potential applications include law enforcement and psychological analysis.

Further information:

University of Oulu

Recent News

25 May 2021

The face recognition imager consumes 10,000 times less energy than a typical camera and processor. CEA-Leti is working with STMicroelectronics on the imager

06 May 2021

The GTOF0503 sensor features a 5µm three-tap iToF pixel, incorporating an array with a resolution of 640 x 480 pixels

30 April 2021

The algorithm can deduce the shape, size and layout of a room by measuring the time it takes for sound from speakers to return to the phone's microphone

20 April 2021

The Kria K26 SOM is built on top of the Zynq UltraScale+ MPSoC architecture. It has 4GB of DDR4 memory and 245 IOs for connecting sensors