Contextual Reasoning for Threat Classification Refinement

This SBIR project will develop a deep neural network for context-based threat classification using features from multiple sensors. We propose to leverage our existing Scenario Knowledge Graph Database to design and encode ontology for threat scenario evolution and identification. Our proposed solution incorporates teacher-student deep networks that are trained using statistical relational learning framework based on Markov Logic Network (MLN) for context-aware target classification and selection for missile defense systems in the presence of countermeasures and clutter in realistic scenarios. The proposed solutions will be built upon the vast experience and expertise of the Novateur Team in the areas of contextual reasoning, deep learning, target classification/recognition, missile defense, radar systems, and multi-sensor data and decision fusion. Approved for Public Release | 21-MDA-11013 (19 Nov 21)