Advancing Target Tracking Systems: Using AI to Recognize Objects in Milliseconds
Intelligence, surveillance, and reconnaissance platforms such as satellites, UAVs, and autonomous systems must process data in milliseconds to transmit decision-making information to the warfighter in real-time. Timing is critical to their mission; the faster they are able to receive actionable intelligence, the quicker they can take action. Riverside Research scientists are pioneering algorithms to make this possible.
Intelligence, surveillance, and reconnaissance platforms such as satellites, UAVs, and autonomous systems must process data in a matter of milliseconds to transmit real-time, decision-making information to the warfighter. Timing is critical to the warfighter’s mission; the faster they are able to receive actionable intelligence, the quicker they can take action.
Novel AI hardware called neuromorphic architectures mimic the speed and processing capabilities of the human brain. They operate with very low size, weight, and power and incorporate third generation neural networks called “spiking neural networks (SNNs),” a very realistic neural simulation that is ideal for processing complex information. In theoretical applications, SNNs have proven useful, but AI scientists have only recently started to develop applications using these networks.
Riverside Research AI scientists are developing algorithms that capitalize on the potential these SNNs have for enabling object detection, tracking, and classification in milliseconds. They are pioneering algorithms that make it possible to use continuous data streams on the order of the speed of the human brain to process incident photons and recognize an object.
Ultimately, these algorithms can be applied to develop next generation intelligence systems that can process image information more efficiently and accurately, saving critical time for the warfighter.