TECHNOLOGY AREAS: Information Systems, Sensors

ACQUISITION PROGRAM: PM Intel, Marine Corps Systems Command, ACAT IV

OBJECTIVE: Develop technologies for an automated capability to recognize and classify entities in video imagery using soft biometrics. Advance understanding of which soft biometrics features/ metadata from entities in video imagery is most useful for subsequent entity identification. Video can include imagery from wide area surveillance cameras both in indoor and outdoor settings. Resultant system should have capability to translate an image to soft biometric metadata about the observed entities. The metadata, once forwarded, should allow a second imager to infer with confidence a second sighting of the same entity. The developed algorithms should also allow for the translation of a description of a person to a similar soft biometric metadata representation, allowing distributed imagers to infer matches. It should be noted that using soft biometrics does not provide absolute identification and verification but is intended to reduce the number of subjects to be investigated.

DESCRIPTION: Soft biometrics are the human characteristics that provide information about the individual but is insufficient to differentiate any two individuals and thus identify an individual reliably and uniquely due to its lack of distinctiveness and permanence. Soft biometrics include information such as gender, eye color, ethnicity, age, height, weight, length of arms/leg, gait and gestures. In order to perform identification or verification, a system totally based on soft biometrics information cannot provide results with a satisfactory matching rate. However they are usually easier to capture from a distance and do not require cooperation from the subject. These soft biometrics traits are descriptors that can be present in human intelligence reports. Soft biometrics can complement and improve the performance of common biometric systems (e.g., signature verification system, face recognition system, fingerprint identification system, iris). The capability sought will allow individual people to be rapidly associated with groups of people having similar soft biometric features. This filter will allow the warfighter to more efficiently process individuals that may be of interest as opposed to individuals with no soft biometric matches to entities of interest.

PHASE I: Conduct research to evaluate the viability of extracting soft biometrics from video imagery. Identify good data sets that can be used for algorithm verification. Develop algorithms that can extract soft biometrics from imagery and record these as metadata in a database. Develop algorithms that can take in soft biometrics descriptors and find matching entities in video imagery. It is expected that performance predictions will be developed during Phase I for all algorithms proposed by the offeror. Submit a report covering the approach, design and results.

PHASE II: Develop a working prototype for the phase I capability and demonstrate its capability against a relevant data set. Deliver and demonstrate the working prototype. Deliver a final report documenting the performance and capability.

PHASE III: Demonstrate that the products developed under phase I and II can be integrated into the Tactical Exploitation Group, the Marine Corps senior Imagery Intelligence (IMINT) system. Provide documents and prototypes to many DoD and contractor test facilities.

PRIVATE SECTOR COMMERCIAL POTENTIAL/DUAL-USE APPLICATIONS: Topic has direct relevance to military operations and civilian law enforcement in that it will develop capability to filter video imagery based on soft biometrics. The topic would enable law enforcement to send descriptions of people of interest to camera networks and be returned possible matches.

REFERENCES:
1. Anil K. Jain, Sarat C. Dass, Karthik Nandakumar , "Soft Biometric Traits for Personal Recognition Systems" Pro. Of the Int. Conf on Biometric Authentication, LNCS 3072, pp 731-738, Hong Kong, July 2004

2. J. Phillips, "Human Identification Technical Challenges", IEEE Int Conf On Image Processing, 2002

3. http://sphericalcube.com/biometrics/biometrics.php

KEYWORDS: Biometrics, soft biometrics, recognition, identification, verification, video analysis

TPOC: Martin Kruger
Phone: (703)696-5349
Fax:
Email: [email protected]
2nd TPOC: Andre des Rosiers
Phone: (202)404-4436
Fax:
Email: [email protected]
3rd TPOC: Thomas McKenna
Phone: (703)696-4503
Fax: (703)696-1212
Email: [email protected]

** TOPIC AUTHOR (TPOC) **

DoD Notice:   Between November 13 and December 9, 2007, you may talk directly with the Topic Author(s) to ask technical questions about the topics. Their contact information is listed above. For reasons of competitive fairness, direct communication between proposers and topic authors is not allowed starting December 10, 2007, when DoD begins accepting proposals for this solicitation.

However, proposers may still submit written questions about solicitation topics through the DoD's SBIR/STTR Interactive Topic Information System (SITIS), in which the questioner and respondent remain anonymous and all questions and answers are posted electronically for general viewing until the solicitation closes. All proposers are advised to monitor SITIS (08.1 Q&A) during the solicitation period for questions and answers, and other significant information, relevant to the SBIR 08.1 topic under which they are proposing.

If you have general questions about DoD SBIR program, please contact the DoD SBIR Help Desk at (866) 724-7457 or email weblink.