Fri, Mar 09, 2018
Artificial Intelligence Would Analyze Video Captured By UAVs
Google has entered into a contract with the U.S. Department of Defense to work on Project Maven, which is designed to use Artificial Intelligence (AI) to identify objects in video and still images captured by drones. And according to Gizmodo, some company employees are not happy that their work is being offered to the Pentagon.
Google says its involvement with Project Maven is not related to use in combat, but some in the tech industry remain unconvinced that such uses would never be considered.
Project Maven, which is also known as Algorithmic Warfare Cross-Functional Team (AWCFT), has been around since April of last year. The DOD spent $7.4 billion on AI-related research in 2017, according to the Wall Street Journal.
The initial charge for the project was to assist in processing the mountains of images and video captured by military drones every day. Human analysts are unable to keep up with the inflow of images and data, according to Greg Allen of the Center for a New American Security. The military has gotten very good at collecting data, but has lagged in ways to make sense of it all.
The Maven program's first objective was using AI to identify vehicles and other objects in up to 38 categories. It also allows the Pentagon to track specific people as they move between different locations.
In a statement to Gizmodo, Google said that it does provided the DOD with TensorFlow APIs, which are used in machine learning applications which help military analysts detect objects in images.
“We have long worked with government agencies to provide technology solutions. This specific project is a pilot with the Department of Defense, to provide open source TensorFlow APIs that can assist in object recognition on unclassified data,” a Google spokesperson said in the statement. “The technology flags images for human review, and is for non-offensive uses only. Military use of machine learning naturally raises valid concerns. We’re actively discussing this important topic internally and with others as we continue to develop policies and safeguards around the development and use of our machine learning technologies.”
(Image from file)
More News
Runway Centerline Lighting Flush centerline lights spaced at 50-foot intervals beginning 75 feet from the landing threshold and extending to within 75 feet of the opposite end of t>[...]
Aero Linx: Air Force Global Strike Command Air Force Global Strike Command, activated August 7, 2009, is a major command with headquarters at Barksdale Air Force Base, Louisiana, i>[...]
Also: Viasat-uAvionix, UL94 Fuel Investigation, AF Materiel Command, NTSB Safety Alert Norges Luftsportforbund chose Aura Aero's little 2-seater in electric trim for their next gli>[...]
Also: USCG Retires MH-65 Dolphins, Irish Aviation Authority, NATCA Warns FAA, Diamond DA42 AD This summer, history enthusiasts will have a unique opportunity to experience World Wa>[...]
Also: Moya Delivery Drone, USMC Drone Pilot, Inversion RAY Reentry Vehicle, RapidFlight UAVOS has recently achieved a significant milestone in public safety and emergency services >[...]