Improving COBRA Performance Using Context-adaptive Algorithms and Algorithm Optimization
Navy SBIR FY2015.1


Sol No.: Navy SBIR FY2015.1
Topic No.: N151-049
Topic Title: Improving COBRA Performance Using Context-adaptive Algorithms and Algorithm Optimization
Proposal No.: N151-049-0241
Firm: Arete Associates
P.O. Box 2607
Winnetka, California 91396-2607
Contact: John Engel
Phone: (520) 571-8660
Abstract: This proposal addresses issues concerning operation of the Coastal Battlefield Reconnaissance and Analysis (COBRA) system in unseen locales and against unseen target types. The COBRA detection algorithms are based on statistical classifiers that are trained using samples from past data collects. As a result, the algorithms may not perform well against unseen situations. Further, once data have been collected in new situations the algorithms may need to be re-optimized, but this is currently a time-consuming effort that requires subject matter expertise. To mitigate these issues Aret� proposes a two-fold approach that aims to improve COBRA detection performance both in previously seen and in new, unseen locales. One aspect of this approach involves developing an offline, stand-alone application called the COBRA Optimization Tool (COBRA OT), which automatically determines the best values of algorithm parameters to provide optimal performance of the algorithms. The second thrust involves creating context-adaptive algorithms by identifying and exploiting information concerning the background settings (contexts) of data collects to improve local detection performance. The two aspects of Aret�'s approach are complementary: context-adaptive algorithms will improve performance in each locale, and COBRA OT will optimize algorithm performance globally across all locales.
Benefits: The successful Phase I and Phase II program will provide improved COBRA algorithm performance across all data collection locales. The resulting validated context-adaptive minefield detection algorithm will significantly enhance the legacy COBRA algorithm. A key metric of the new algorithm is robustness - it will meet COBRA KPP requirements automatically at each individual locale. This program will also result in a stand-alone optimization software tool, COBRA OT, which will enable automatic optimization of the COBRA algorithm processing chain as data in new environments are acquired and/or as algorithms are updated. COBRA OT will be delivered to the COBRA SSA. Future applications of the context-adaptive algorithms will benefit other COBRA missions, including Automated Obstacle Detection (AOD), Near-Surface Naval Mine Detection (NSNMD), and Tactical Littoral Sensing (TLS). The environment-adaptive nature of these algorithms will also benefit future COBRA Block II applications involving new operational environments, including surf zone mine detection and day/night operation. The technology developed under this program will have multiple commercialization applications. The first application is the insertion of a new algorithm on the COBRA Real-Time Processor (RTP) for the COBRA Program, which is a DoD Program of Record (PoR). The new algorithm will be inserted through a Phase III SBIR contract. Additionally, Aret� will deliver an Optimization Tool to the COBRA Software Support Activity (SSA). The Optimization Tool will decrease algorithm optimization timelines from many weeks to a few days or less and will help automate the process. Aret� will provide support and updates for the Optimization Tool under a Phase III contract. The environmentally adaptive algorithms and optimization tool have potential application to many Navy programs with ATR applications. These techniques and tools will be marketed to programs such as ALMDS, AMNS, AQS-20, and MK-18. In addition to DoD applications, the technology developed under this program has value in commercial applications such as autonomous drone-based search for law enforcement.

Return