TRACE – Target Recognition with the Assistance of Artificial Intelligence

Intelligent Automation Inc. (IAI) proposes the design and implementation of the Target Recognition with the Assistance of Artificial Intelligence (TRACE) system that incorporates classical model-based target classification and identification approaches with data-driven machine learning solutions to improve the target classification accuracy. The outcome of the TRACE systems is a set of machine learning and sensor data fusion algorithms. The key innovation of TRACE is the design of an ensemble of supervised learning architectures where a dedicated neural network is trained for each input sensor system; sensor data fusion algorithms which performs decision-level fusion by incorporating the results of model-based approach with neural network architectures using a probabilistic model; and unsupervised learning architectures that extract features for unknown vessel types thereby facilitating their classification. The performance of TRACE will be extensively tested using both the synthetic data that we generate and publicly available sensor system input data sources.