Remote Thermal Imaging for Reliable Deception Detection

Thermal imaging of the face and head for the purpose of deception detection has shown significant promise in recent studies. This technique could be used to enhance polygraph examinations and, because it is noninvasive, could also potentially be used for mass screening at critical security checkpoints. The primary technical obstacles that must be overcome in order to realize a system suitable for these applications are: (1) achievement of adequate thermal resolution to detect the signals of interest, (2) localization of the appropriate target area at a distance, and (3) realization of automated signal processing software. Barron Associates, Inc. (BAI) has teamed with Mayo Clinic to propose a research program that will enhance the understanding of the physiology of deception and will produce a system targeted for employment in mass screening applications. The Mayo Clinic will perform clinical experiments using its system for thermal imaging of the face (STIF) to capture the transient thermal responses associated with deception. BAI will implement invariant feature extraction algorithms, apply these algorithms to the captured thermal image sequences, and use these invariant features as input to its GNOSIS neural network synthesis software to synthesize automated classification networks that can be used for real-time deception detection.