TECHNOLOGY AREAS: Information Systems, Sensors, Battlespace, Human Systems

ACQUISITION PROGRAM: PEO LMW, PMS-495, Organic Post Mission Analysis (non-ACAT)

The technology within this topic is restricted under the International Traffic in Arms Regulation (ITAR), which controls the export and import of defense-related material and services. Offerors must disclose any proposed use of foreign nationals, their country of origin, and what tasks each would accomplish in the statement of work in accordance with section 3.5.b.(7) of the solicitation.

OBJECTIVE: Develop advance recognition algorithms and techniques capable of robust in situ adaptation based on class labels, label confidence, and / or environmental context. The goal is to achieve a substantial reduction in false alarm rates for recognizing mines and other underwater targets of interest.

DESCRIPTION: The task of recognizing objects underwater is highly challenging, and human operators typically outperform computer algorithms in moderate and difficult environments. This is due to the high variability in environmental conditions and human proficiency in incorporating in situ information and context. Most recognition algorithms in use to date are globally pre-trained across many environments or target types and employ little or no in situ adaptation. Current research in techniques for in situ incorporation of class labels (e.g., concept drift or ensemble learning) show promise but lack robustness to noisy or spurious data. Further research is also required to develop robust techniques capable of recognizing environmental context and adapting appropriately. To this end, ONR is seeking recognition algorithms and techniques capable of 1) robustly adapting the parameters of or algorithms for detection, feature extraction, feature selection, and classification based on in situ class labels, label confidence, or environmental type or context, 2) characterizing or learning which environments warrant slightly modified algorithms vice fundamentally different detection and feature extraction approaches, and / or 3) robust learning across sensor types, environments, and target classes.

PHASE I: Development of overall concept, detailed description of how algorithms will be adapted, on what information the adaptation is based, and a simple demonstration on government provided data where only a limited portion of the algorithms need be adapted (e.g., adapting the feature extraction process)

PHASE II: Extend proof-of-concept algorithms from Phase I to robustly and substantially adapt in a laboratory environment given a government provided dataset. This will likely require adaptation of multiple facets of the recognition technique (e.g., classification boundaries, techniques, detection techniques, feature sets, etc).

PHASE III: Extend algorithms to be robust and fault tolerant to a full spectrum, government provided dataset. Work with government labs and contractors to integrate into program of record.

PRIVATE SECTOR COMMERCIAL POTENTIAL/DUAL-USE APPLICATIONS: These techniques are applicable to all industries requiring underwater surveys, searches, or mapping including petroleum, utilities, and geology.

REFERENCES:
1. J. C. Hyland, G. J. Dobeck, "Sea Mine Detection and Classification Using Side-Looking Sonar," Proceedings of SPIE''''95, Vol. 2496, pp. 442-453, Orlando, Florida, 17-21 April 1995.

2. G. J. Dobeck, "Fusing Sonar Images for Mine Detection and Classification," Proceedings of SPIE''''99, Vol. 3710, pp. 602-614, Orlando, Florida, 5-9 April 1999.

3. G. Widmer and M. Kubat, "Learning in the presence of concept drift and hidden contexts," Machine Learning, vol. 23, pp. 69�101, 1996.

4. H. Wang, W. Fan, P. S. Yu, and J. Han, "Mining concept-drifting data streams using ensemble classifiers," KDD �03: Proceedings of the ninth ACM SIGKDD international conference on Knowledge discovery and data mining, New York, NY, USA, 2003, pp. 226�235, ACM Press.

KEYWORDS: Adaptation; automatic target recognition; autonomous systems; classification; machine learning; mine countermeasures; sea mines

** TOPIC AUTHOR (TPOC) **

DoD Notice:   Between November 12 and December 7, 2008, you may talk directly with the Topic Authors 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 8, 2008, 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 (09.1 Q&A) during the solicitation period for questions and answers, and other significant information, relevant to the SBIR 09.1 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.