A University of Notre Dame professor of computer science and engineering says listening to military and national security experts discuss the need for a better way to ID bomb makers in the Middle East inspired him to join forces with colleagues designing a Questionable Observer Detector.
Notre Dame's Kevin Bowyer has teamed with colleague Patrick Flynn and doctoral student Jeremiah Barr to research a biometrics-based system for identifying criminals, including bomb makers, who return to the scene of the crime. Their work on biometrics over the years has included analysis of everything from facial photographs to face thermograms to iris images to videos of the human gait.
The Questionable Observer Detector is designed to spot people who show up repeatedly in video taken at crime scenes, even when the video isn't of great quality. The researchers create what they call "face tracks" for anyone appearing in a video and then seek matching face tracks in other videos.
The researchers acknowledge possible civil liberties concerns, but they argue the technology should enable law enforcement to identify suspects based on stronger evidence than is currently available.
The researchers explain their work further in this video from Notre Dame: