You are here

Advanced Artificial Intelligence/Machine Learning Techniques for Automated Target Recognition (ATR) Using Small/Reduced Data Sets


OUSD (R&E) CRITICAL TECHNOLOGY AREA(S): Integrated Sensing and Cyber; Trusted AI and Autonomy


The technology within this topic is restricted under the International Traffic in Arms Regulation (ITAR), 22 CFR Parts 120-130, which controls the export and import of defense-related material and services, including export of sensitive technical data, or the Export Administration Regulation (EAR), 15 CFR Parts 730-774, which controls dual use items. Offerors must disclose any proposed use of foreign nationals (FNs), their country(ies) of origin, the type of visa or work permit possessed, and the statement of work (SOW) tasks intended for accomplishment by the FN(s) in accordance with the Announcement. Offerors are advised foreign nationals proposed to perform on this topic may be restricted due to the technical data under US Export Control Laws.


OBJECTIVE: Develop an innovative automatic target recognition (ATR) capability that leverages state-of-the-art Artificial Intelligence/Machine Learning (AI/ML) technology to support naval mine countermeasure (MCM) operations using reduced data sets.


DESCRIPTION: ATR is the ability for a system or algorithm to recognize and identify targets, objects of interest or threats based on data obtained from sensors. In Navy mine countermeasure (MCM) operations, sensors collect data to identify and localize targets of interest in marine environments.

The Navy is interested in developing state-of-the art AI/ML ATR processing algorithms, or techniques to facilitate target detection and identification using smaller data sets to train the algorithms and perform ATR. The Navy’s existing Minehunting systems collect data using forward-looking sonar, a pair of side-scan sonars, and a volume search sonar. Identification and localization of underwater objects is challenged by both a reliance on large, curated data from the onboard sensors that are needed to train and perform ATR and the amount of time required to conduct ATR operations. Current MCM ATR algorithms require large amounts of data (over 200 hours of acoustic video and 1,000-2,000 target images) to train the algorithms. This training data is quite costly to obtain because it must be collected in a variety of representative operational environments.


The proposed solution should demonstrate reduction in the amount of data required to train algorithms by an order-of-magnitude smaller without degradation to identification performance (Pid) and no increase in the Probability of false alarms (Pfa). If possible, the solution should incorporate advanced ML techniques such as One Shot, Multi Shot, Few Shot etc. as well as others that yield the benefits sought.

The ATR will be initially integrated into the Navy’s Generalized ATR (GATR) framework to improve detection and classification performance. The capability could eventually be integrated into a towed body to support in-stride ATR.


Work produced in Phase II may become classified. Note: The prospective contractor(s) must be U.S. owned and operated with no foreign influence as defined by 32 U.S.C. § 2004.20 et seq., National Industrial Security Program Executive Agent and Operating Manual, unless acceptable mitigating procedures can and have been implemented and approved by the Defense Counterintelligence and Security Agency (DCSA) formerly Defense Security Service (DSS). The selected contractor must be able to acquire and maintain a secret level facility and Personnel Security Clearances. This will allow contractor personnel to perform on advanced phases of this project as set forth by DCSA and NAVSEA in order to gain access to classified information pertaining to the national defense of the United States and its allies; this will be an inherent requirement. The selected company will be required to safeguard classified material during the advanced phases of this contract IAW the National Industrial Security Program Operating Manual (NISPOM), which can be found at Title 32, Part 2004.20 of the Code of Federal Regulations. Reference: National Industrial Security Program Executive Agent and Operating Manual (NISP), 32 U.S.C. § 2004.20 et seq. (1993).


PHASE I: Develop a concept to facilitate target identification using smaller data sets that meets the requirements described above. Demonstrate the feasibility of the concept in meeting Navy needs and establish that the concept can be feasibly developed into a useful product for the Navy. Feasibility will be established by testing and analytical modeling.


The Phase I Option, if exercised, will include the initial design specifications and capabilities description to build a prototype solution in Phase II.


PHASE II: Develop and deliver a prototype for evaluation as appropriate. The prototype will be evaluated to determine its capability in meeting the performance goals defined in the Phase II SOW and the Navy requirements for the algorithms. Demonstrate performance across a broad set of Government -Furnished Information (GFI) data. Performance will be validated against Government-provided target truth. Prepare a Phase III development plan to transition the technology to Navy use.


It is possible that the work under this effort will be classified under Phase II (see Description section for details).


PHASE III DUAL USE APPLICATIONS: Produce and support a final prototype that will be integrated into developmental and operational frameworks used by the AN/AQS-20 Mine Hunting Sonar Post Mission Analysis (PMA)/ Generalized ATR (GATR) system. Additionally, AI/ML algorithms developed may be inserted onboard the AN/AQS-20 Mine Hunting Sonar towed body. Due to the nature of the effort coupled with the anticipated implementation of DEVSECOPS, technology insertions may also be accelerated and/or incrementally introduced into various other MCM sensors (e.g., Mk18 FOS, AQS-24, Barracuda, etc.).


Technology developed under this effort is applicable to any domain that requires subsea platform autonomy such as subsea oil and gas pipeline inspection.



  1. Koch, G.R. (2015). Siamese Neural Networks for One-Shot Image Recognition. Department of Computer Science, University of Toronto. Toronto, Ontario, Canada, 2015.
  2. Tiu, E., Talius, E., Patel, P. et al. Expert-level detection of pathologies from unannotated chest X-ray images via self-supervised learning. Nat. Biomed. Eng 6, 1399–1406 (2022).
  3. Wang, Yaqing, Yao, Quanming, Kwok, James T. and Ni, Lionel M.(2020). Generalizing from a Few Examples: A Survey on Few-shot Learning. ACM Comput. Surveys, , 53(3).


KEYWORDS: Artificial intelligence; machine learning; mine countermeasures; acoustic sensor; automatic target recognition; few-shot learning

US Flag An Official Website of the United States Government