U.S. Army researchers have developed a new method that allows neural networks to be more confident in judging potential threats in battlefield environments as part of their artificial intelligence study.
The Army Research Laboratory (ARL) and Army Combat Capabilities (ACC) Development Command worked with university researchers from the Internet of Battlefield Things Collaborative Research Alliance (IoBT CRA) to classify sources of uncertainty, assess frameworks to represent uncertainty and developed platforms to manage uncertainty within systems, the service said Monday.
The researchers derived insights from uncertainty management approaches and turned them into a workflow that maximizes effectiveness in realizing mission goals amid the presence of uncertainty in data inputs. The process enabled the neural networks to be more confident in judging threats in hostile environments.
“A key component of improving automation is to improve machine confidence in understanding its environment so that the machine can exercise ‘good judgment,’” said Maggie Wigness, Army researcher and deputy collaborative alliance manager of the IoBT CRA.
