WASHINGTON – The National Geospatial-Intelligence Agency has selected a team of commercial and academic partners to build an artificial intelligence system with synthetic data, which will further help the agency determine how it builds machine learning algorithms moving forward.
Orbital Insight was issued a Phase II Small Business Innovation Research contract by the NGA, the company announced. Dec. 16. It will collaborate with Rendered.ai and the University of California, Berkeley, to develop a computer vision model
As the organization charged with analyzing satellite imagery for the intelligence community, NGA has put increased emphasis on using AI for its mission. The agency sees human-machine pairing as critical for its success, with machine learning algorithms taking over the rote task of processing the torrent of satellite data to find potential intelligence and freeing up human operators to do more high level analysis and tasks.
“There is an influx in imagery and data that humans can’t analyze with eyesight alone,” Orbital Insight CEO Kevin O’Brien said in a statement. “While our national security relies on this data, computer vision can help provide the right answers. We’re honored to have been awarded this Phase II contract and continue our partnership with the NGA. The results of this project will be instrumental for the defense intelligence community.”
To build effective AI, engineers need to feed massive amounts of training data — images of the types of objects they want it to find — to machine learning algorithms so they can automatically find those objects when presented with new images. While companies use real world data to train these algorithms, there simply isn’t enough. Real world data is often supplemented with synthetic data, entirely artificial but designed to look like the real world data the machine is being built to work with.
Most algorithms used by the government are trained on both real world images and synthetic images, but this effort will solely use synthetic data. Orbital Insight says the project will improve efforts to use mixed training data to build AI tools.
Nathan Strout covers space, unmanned and intelligence systems for C4ISRNET.
