Micro Expressions in Detecting Deceptive Behavior

Deceptive behavior is a common phenomenon in interpersonal interactions. Although people have more or less experienced deception, whether as deceivers or deceived, it is well known that lies are difficult to detect—even for experienced experts in law enforcement. Despite this, many methods for detecting lies have emerged, with the polygraph being one of the more common methods currently used by law enforcement. The polygraph assesses physiological responses caused by the sympathetic nervous system, such as blood pressure, pulse, breathing, and skin conductivity, to determine if a subject is lying. However, the polygraph process is invasive, and subjects may employ various countermeasures to conceal their true thoughts.

Recent research has identified another potential indicator of deception—micro expressions. Micro expressions are rapid and brief facial expressions that occur when a person tries to conceal their true emotions. They are characterized by their extremely short duration, low intensity, and usually localized movements. Unlike polygraphs, micro expressions can be captured using unobtrusive cameras during interactions, so subjects may not be aware that they are being observed.

Previous studies have established facial expression databases that include both macro and micro expressions from long video sequences. Part A contains 87 long videos, featuring both expressions and micro expressions. Part B includes 300 cropped expression samples and 57 micro expression samples. Emotional labels are based on Action Units (AUs), self-reported emotions for each facial movement, and the emotional types of the videos that triggered the emotions. These expression samples were collected in the same experimental context, and the database allows future researchers to develop more efficient algorithms for extracting and distinguishing micro expression features from long videos. While automatic detection of micro expressions is still under development, advancements in this technology could enable its widespread application in various contexts for detecting deception in the future.

Ref: https://ieeexplore.ieee.org/document/7820164

Photo from CAS(ME)2: A Database for Spontaneous Macro-Expression and Micro-Expression Spotting and Recognition