Authors : Harith H. Thannoon, Wissam H. Ali and Ivan A. Hashim
Abstract:
Human face is a wealthy source of information which provides reliable cues to deception. The Deceptive Detection Systems (DDSs) through the identification of facial expression are non-invasive, mobile and cost effective. In this study, the DDS is designed depending on Facial Action Coding System (FACS) to extract the facial features. The main idea of this FACS is coded the facial muscles movements using Action Units (AUs). Each AU represented the movement of certain facial muscle. The proposed system discriminates lying subjects from the innocent one based on presence or absent the facial AUs. Eight AUs are used as potential indicators for deception that incorporated into a single facial behavior pattern vector. Database that used to validate the proposed system are collected from 43 subjects (20 males, 23 females) most of them between ages 18-25. The number of video clips that obtained from collected database after editing was 400 video clips. Virtual Generalizing Random Access Memory Weightless Neural Network (VG-RAM WNN) classifier is used to make decision in last stage of DDS. The proposed DDS was tested three times, accuracy 84, 85 and 90% of spotting liars are achieved when using both genders, only female and only male participants respectively. The VG-RAM classifier was built by FPGA Model using Xilinx system generator and implemented on Spartan-3A 700 A Kit.
Harith H. Thannoon, Wissam H. Ali and Ivan A. Hashim, 2019. Design and Implementation of Deception Detection System Based on Reliable Facial Expression. Journal of Engineering and Applied Sciences, 14: 5002-5011.