Home / News / Col. Drew Cukor: DoD Aims to Extract Objects From Imagery With New Algorithms

Col. Drew Cukor: DoD Aims to Extract Objects From Imagery With New Algorithms

The Defense Department plans to deploy new computer algorithms designed to extract objects of interest from moving or still imagery by the end of the year, DoD News reported Friday.

U.S. Marine Corps Col. Drew Cukor, head of the Algorithmic Warfare Cross-Function Team in DoD’s Office of the Undersecretary of Defense for Intelligence, said at an event hosted by Defense One that the new algorithms were developed as part of the Project Maven effort that seeks to help DoD analysts break down huge amounts of data.

“Eventually we hope that one analyst will be able to do twice as much work, potentially three times as much, as they’re doing now,” Cukor added.

Project Maven is focused on computer vision, a branch of machine and deep learning that involves the autonomous extraction of objects from videos and still imagery.

DoD will initially use computer vision on 38 types of objects that the department needs to detect to support missions, such as the fight against the Islamic State militant group, Cukor noted.

He added that DoD analysts and engineers will triage and label the department’s data over the next few months to prepare for machine learning.

DoD also looks to buy additional computational power technology such as graphic processing units designed to support training of machine-learning algorithms as well as launch a competitive procurement process for an algorithmic development contract.

Check Also

AFRL Seeks Health Monitoring, Warfighter Dev’t Concepts

The Air Force Research Laboratory is looking for concepts to speed up human-monitoring research and development initiatives for warfighters. The laboratory is in need of prototypes to support performance enhancement, health monitoring and diagnostics systems, the Wright-Patterson Air Force Base said Thursday.

Leave a Reply

Your email address will not be published. Required fields are marked *