Description of Invention
Traditionally, a polygraph machine is used to measure the heartbeat rate, skin conductivity and respiration rate to observe any changes that relates to lying. However, a polygraph test might not be accurate. A modern way to overcome the problem is by observing the facial micro-expression that is unintentionally produced by the facial muscles. Facial Action Coding System (FACS) is a system that categorises different facial muscles into the Action Unit (AU), developed by Dr. Paul Ekman. Many researches have been done, including machine learning (ML), to detect lying behaviours based on the features of AU. However, most of these models require huge computer processing power, and it is not feasible for real-time applications. It also dependent on the extraction of the AU’s features using specific software such as OpenFace 1.0. Hence, this project presents a faster-processing approach to detect possible lying behaviours based on the extracted facial micro-expressions using OpenFace 2.0 and deep learning.