Researcher to develop AI system to track terrorists

Computer and behavioral scientists at the University at Buffalo are teaming up to create an artificially intelligent program prototype to track and analyze different behavioral biometrics such as faces, voices, and body movements to pinpoint individuals who are likely about to commit a terrorist act. Dr. Venu Govindaraju, professor of computer science and engineering, and Dr. Mark Frank, associate professor of communications, are the principle researchers on the UB project. They intend for the computer program to be used in security settings to identify terrorists before they carry out their attacks. The prototype will be able to examine a video in a number of different security settings based on the behavioral biometrics that are witnessed during interrogations. It will be important for the computer to be able “learn” from the subjects being interrogated because behavioral science research consistently demonstrates that the behavioral clues that indicate a person is lying are relative from person to person. It will be crucial for the computer program to be able to track an individual’s behaviors and create a behavioral baseline specific only to that person.

Frank argues that while no behavior accurately indicates that a person is lying, they do predict emotions or thoughts. Becoming more aware of emotions and thoughts will help a security officer decide who to keep an eye on. Frank and Govindaraju believe that the UB system will provide better security by providing a scientific basis for further security screening rather than just random screening. Security guards feel tired and feel bias, but the computer program will feel neither of the above. The UB project is the first to combine the work of computer scientists with behavioral scientists to develop a more accurate security detection system. Frank, founder of the Communication Science Center at UB, has dedicated his career to researching human body language and its indication of emotions and telling the truth. Frank’s work has already been recognized and used by security officials all over the world. Similarly, Govindaraju’s work has focused on problems in artificial intelligence recognizing patterns. He founded the Center for Unified Biometrics and Sensors in 2003 where work is being done on developing technologies which combine different biometrics detections. Together, the two are creating new algorithms to show patterns that will help to pinpoint a terrorist. They expect the prototype of the full system to be ready in a few years.