Objective: To develop new state-of-the-art information fusion algorithms focused on collaborative ATA/ATR in which multiple agents with diverse sensors are combined to improve multi-target tracking in near-real-time (seconds).

Environment: The algorithms must identify, classify, and track targets in an environment that includes multiple kinds of targets, confounders, and natural clutter. The information sources will include diverse sensors (RF/EO/IR/magnetic/acoustic) obtained from airborne platforms that are not precisely geolocated with relation to each other. Targeting decisions must be made within seconds to at most a minute in a decentralized architecture. The computing, communication, and memory storage limitations of typical loitering munitions and UAVs will be important, also.

The focus is on target identification and tracking, rather than on guidance and coordination of the swarming munitions.

Approach: The algorithmic approaches will probably combine some state-modeling (e.g., hypergraphs, or variants of Markov Decision Processes) along with a combination of various kinds of strong and weak AI/Machine Learning. Topics such as assigning value to incremental knowledge through calculating Entropy and context-aware fusion are also of high interest.

Dong, Y., Feng, S. and Tai, L., 2021. Three-dimensional target tracking strategy based on guidance laws and optimal information fusion. Transactions of the Institute of Measurement and Control, 43(7), pp.1560-1570.

Pan, Q., Hu, Y., Lan, H., Sun, S., Wang, Z. and Yang, F., 2019. Information fusion progress: Joint optimization based on variational Bayesian theory. Zidonghua Xuebao/Acta Automatica Sinica, 45(7), pp.1207-1223.

Jin, X.B., Sun, S., Wei, H. and Yang, F.B. eds., 2018. Advances in multi-sensor information fusion: Theory and applications 2017. MDPI.

Keywords:

Information Fusion; Sensor Fusion; Target Acquisition; Hierarchical Learning; Machine Learning

Citizenship:  Open to U.S. citizens

Level:  Open to Postdoctoral and Senior applicants