DoD SBIR 23.3 BAA

Active
No
Status
Closed
Release Date
August 23rd, 2023
Open Date
September 20th, 2023
Due Date(s)
October 18th, 2023
Close Date
October 18th, 2023
Topic No.
AF233-0015

Topic

Multi-int Multi-look 3D features Image Fusion with Machine Learning

Agency

Department of DefenseN/A

Program

Type: SBIRPhase: BOTHYear: 2023

Summary

The Department of Defense (DoD) is seeking proposals for the topic "Multi-int Multi-look 3D features Image Fusion with Machine Learning" as part of the SBIR 23.3 BAA. The objective of this solicitation is to develop novel, computationally efficient multi-int and multi-look fusion algorithms to improve the detection, classification, and recognition of uncooperative low resolution and occluded objects in urban and complex terrains. The technology will leverage modern sensors and machine learning algorithms to achieve these goals. The project will be conducted in three phases. In Phase I, state-of-the-art solutions in feature extraction and object-level probability of detection/identification methods will be investigated. An architecture design concept for the Phase II algorithms will be developed, along with a complete description of the proposed algorithms. The Phase I report should include a demonstration of electro-optic, infrared, or radar data processing and low-level feature fusion from image data. In Phase II, a functional algorithmic suite and operator interface will be developed and demonstrated using realistic sensor data. Measures of performance established in Phase I will be validated, and the project will be extended to include additional sensor modalities and operating conditions. The final report should document progress made and include requirements to sensors. Source code and data sets for all techniques developed under the contract will be delivered. In Phase III, the developed system could be used in a broad range of military and civilian security applications, such as military operations in urban and complex terrain, search and rescue, firefighting, drug interdiction, law enforcement, and counter-terrorism operations. Commercial applications include remote sensing, industrial development and operations, traffic analysis, and environmental monitoring. The project is restricted under the International Traffic in Arms Regulation (ITAR) or the Export Administration Regulation (EAR), and offerors must disclose any proposed use of foreign nationals. The solicitation is currently closed, and more information can be found on the DoD SBIR website.

Description

OUSD (R&E) CRITICAL TECHNOLOGY AREA(S): Trusted AI and Autonomy;Advanced Computing and Software

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: To develop novel, more computationally efficient multi-int and multi-look fusion algorithms to improve detection, classification, and recognition of uncooperative low resolution and occluded objects in urban and other complex terrains using advantages and capabilities of modern sensors and machine learning algorithms.

DESCRIPTION: To develop novel, more computationally efficient multi-int and multi-look fusion algorithms to improve detection, classification, and recognition of uncooperative low resolution and occluded objects in urban and other complex terrains using advantages and capabilities of modern sensors and machine learning algorithms.

PHASE I: Investigate state of the art solutions in simple feature extraction and object level probability of Detection/identification methods that reduce the need for human oversight and the computational complexities of the fusion methods. Develop an architecture design concept of the Phase II detection/classification algorithms. The Phase I report should provide a complete description of the proposed algorithms that includes demonstration of electro-optic, infrared, or radar data processing and low level feature fusion from image data. It is desirable that algorithms be computationally efficient to provide near real-time capabilities. Any data needs and assumptions required by the concept should be clearly outlined and explained. Source code of the final demo is a required deliverable.

PHASE II: Develop and demonstrate a functional algorithmic suite and operator interface using realistic electro-optic, infrared, or radar imagery sensor data. Validate measures of performance established in Phase I. Other tasks include documenting and delivering a report, interim and final source code, including all users’ needs assessments, methodologies, algorithms, and any data structures or software products necessary to support transition of the work to Air Force applications. Imagery and other multi-int data may be provided to the awardees. Extend Phase I approach to include additional sensor modalities and operating conditions. Sensor operating constraints shall also be addressed. The final report should document progress made and include requirements to sensors (range, spectra, timing, minimal amount of multi-look angles/images, and other operational constraints). Final delivery should include source code and data sets for all techniques developed under the contract. To streamline transition of the Phase II products to AF applications, a business model for compensation of the developers’ SBIR data rights must be provided.

PHASE III DUAL USE APPLICATIONS: This system could be used in a broad range of military and civilian security applications where real-time information fusion and target detection/recognition is required: for example, in military operations in urban and complex terrain, in search and rescue, firefighting, drug interdiction, law enforcement, counter terrorism operations in urban structures, border tunnels, industrial facilities, etc. Commercial Application: Technologies developed under this effort can be applied to remote sensing, industrial development and operations, traffic analysis, and environmental monitoring.

REFERENCES:

  1. Igor Ternovskiy, “Scene Understanding Based on Mapping Singularities and New Primitives Generation” 2005 IEEE International Conference on Integration of Knowledge Intensive Multi-Agent Systems KIMAS’05 Boston MA, USA 18 – 21 April, 2005 Workshop on Sapient Systems Editors: Rene V. Mayorga & Leonid I. Perlovsky
  2. I. Ternovskiy, et. al., "Is catastrophe analysis the basis for visual perception?," chapter in Three-Dimensional Holographic Imaging, C.J. Kuo, M.H. Tsai (Eds.), John Wiley and Sons, Inc., 2002. ISBN: 0-471-35894-0
  3. I.V. Ternovskiy, T.P. Jannson, "Rotational invariant visual object extraction and understanding," Proc. SPIE, Signal Processing, Sensor Fusion, and Target Recognition IX, I. Kadar (Ed.), Vol. 4052, pp. 85-93, August 2000.
  4. T.P. Jannson, I.V. Ternovskiy, "Data reduction for multispectral and hyperspectral imagery based on application of catastrophe theory," Proc. SPIE, Applications and Science of Neural Networks, Fuzzy Systems, and Evolutionary Computation III, B. Bosacchi, D.B. Fogel, J.C. Bezdek (Eds.), Vol. 4120, pp. 110-119, October 2000.

KEYWORDS: Machine Learning; electro-optic, infrared, radar sensors; multi-phenomenology fusion; ISR; multi-look