The next generation of
systems is here -- voice recognition, iris and retina scans, facial recognition coupled with DNA analysis. And the FBI is building it.
In addition to those biometrics, the Federal Bureau of Investigation's Next Generation Identification (NGI) program will also include enhanced fingerprint identifying -- rolled and latent finger and palm prints.
A pilot NGI program is under way, and includes the futurist match-up of head shots, taken from crowd images, security systems, or even social networking sites, with known faces in a database. Additionally, the bureau is developing a database of scars and tattoos, which could also be used for visual match-ups.
Tests conducted two years ago indicated that, using the best available algorithms, a face can be successfully matched 92 percent of the time using a pool of 1.6 million mugshots.
Additionally, developed at Carnegie Mellon can take data from front and side views, create a 3D model, and then rotate the model as much as 70 degrees to create a 2D match with the image in the crowd or security system footage. Researchers at Carnegie Mellon have said that computer-based facial recognition, in certain conditions, can now be as good or even better than human facial recognition.
This is not the first time the FBI has collected this kind of data, but it is the first time that such a system is being tested and rolled out nationally. Full implementation of the $1 billion project is expected by 2014.
'Privacy and Civil Liberties'
In July, the FBI's Jerome Pender, deputy assistant director of the Criminal Justice Information Services Division, told a subcommittee of the Senate Judiciary Committee that Michigan has successfully conducted an end-to-end Facial Recognition Pilot test, and has begun to submit facial recognition searches to his division.
Agreements to participate in the Facial Recognition Pilot system have been completed with Hawaii and Maryland. South Carolina, Ohio and New Mexico are reviewing the agreements.
Pender told the Senate that the searchable photo database, which currently contains 12.8 million searchable frontal photos, only contains images of known criminals. His testimony to the Senate was entitled "What Facial Recognition Technology Means for Privacy and Civil Liberties," which is certain to become a much-discussed topic.
NGI is being implemented in several stages. The first stage, which deployed Advanced Technology Workstations, was completed in March 2010, and the second stage, to increase the accuracy of fingerprint searches to 99.6 percent, was finished in February of last year.
A Repository of Individuals of Special Concern, the third stage, was completed in summer 2011, and it also provided fingerprint identification operations.
The next three stages are in progress. They create Palm Print Search and Latent Print Searching, the establishment of a National Palm Print System, and the ability to search scars, marks and tattoos. Stages to implement an iris recognition pilot and a full facial search system will begin next year.