News

FBI's Facial Recognition Program: Better Security Through Biometrics

Michael Endler
Associate Editor, InformationWeek.com

The FBI's facial recognition technology is a boon for law enforcement--and perhaps soon for enterprise and consumer security as well.

The FBI is moving ahead with a nationwide facial recognition program scheduled to be fully deployed by 2014, according to New Scientist and testimony delivered to the Senate in July. The program could lead to faster, more efficient law enforcement--but nabbing crooks after a crime is only part of the appeal. The technology also foreshadows upcoming security enhancements that will stop many offenses before they start, including several that plague businesses.

The new tools are part of the FBI's $1 billion Next Generation Identification (NGI) program, a surveillance initiative built around biometric data.


More Government Insights

Webcasts

More >>

White Papers

More >>

Reports

More >>

This data involves more than facial-recognition tools. Originally conceived to replace the bureau's aging fingerprint identification system, NGI also employs a 10-point fingerprint matching process that is 99% accurate. Other capabilities include the ability to deduce identities from palm prints, tattoos, and potentially even DNA.

[ For more on the FBI's biometric ID program, see FBI To Add Tattoos To Biometric ID Capabilities. ]

Some of these tools won't be widely deployed until NGI is fully operational in summer 2014, but the facial recognition is slowly proliferating. Michigan initiated a beta rollout in February, and at least 10 additional states have either begun testing or expressed interest.

The FBI most recently disclosed details about the pilot program when the bureau's Jerome Pender tesitfied before the Senate in July. He said that NGI's facial recognition tools can compare a query image to a database of 12.8 million mug shots. Such a large database should facilitate easier tracking of suspects who flee across jurisdictions, and research suggests the effects could be dramatic; 2010 tests found that facial recognition tools correctly identified individuals from a pool of 1.6 million mug shots with 92% accuracy.

Newer versions could be even better. Researchers at Carnegie Mellon have developed algorithms that use 3-D modeling to more accurately divine identities from faces, and Alessandro Acquisti, a professor at the university, told the Senate in July that face detection is mature enough for primetime.

Acquisti also expressed caution about the technology's power. Civil libertarians are concerned the technology represents Big Brother as much as big data. They cite, among other things, the FBI's suggestion that NGI could be used to track individuals within crowds. The FBI has taken steps to ensure innocent citizens are not targeted for surveillance, however; Pender told the Senate that query images obtained through social networking sites, surveillance cameras, and similar sources "are not used to populate the national repository."

Outside the government, biometric tech has a mixed record. Facebook inadvertently triggered controversy when it integrated facial-recognition technology into its photo-tagging function. And UPEK fingerprint readers were shown in August to suffer from a vulnerability that could expose passwords.

Other developments have been more auspicious, however. Saratoga Hospital, in Saratoga Springs, NY, used biometric technology provided by DigitalPersona Inc. to more efficiently and securely verify access to confidential records. In the consumer realm, Apple's July acquisition of fingerprint security company AuthenTec suggests biometrics may headline future iOS and OS X enhancements.

Page 2: Trickle Down Security
 1 | 2  | Next Page » 

Related Reading


Informationweek Discussions

Start the Discussion


InformationWeek encourages readers to engage in spirited, healthy debate, including taking us to task. However, InformationWeek moderates all comments posted to our site, and reserves the right to modify or remove any content that it determines to be derogatory, offensive, inflammatory, vulgar, irrelevant/off-topic, racist or obvious marketing/SPAM. InformationWeek further reserves the right to disable the profile of any commenter participating in said activities.

Disqus Tips To upload an avatar photo, first complete your Disqus profile. | View the list of supported HTML tags you can use to style comments. | Please read our commenting policy.
Subscribe to RSS

Resource Links