Page 10 - GCN, April/May 2018

\[BrieFing\]
Blockchain, automation and learning to love your legacy systems
Facial recognition in the dark
BY MATT LEONARD
A conceptual illustration of thermal-to- visible synthesis for interoperability with existing visible-based facial recognition systems.
Facial recognition generally requires well-lit photos to make a match, but warfighters often need to identify individuals from images taken by thermal cameras in low-light or nighttime conditions. With the help of artificial intelligence, scientists at the Army Research Lab have developed
a way to enhance images from night- vision cameras so that they can be matched to conventional images in existing biometric face databases or watch lists.
The approach creates a visible facial image that has been synthesized from multiple regions of a thermal image.
By combining features from across the entire face with those of specific regions — such as the eyes, nose and mouth — the researchers were able to synthesize a refined image that could be run against facial images in a database.
Making a match between the visible and thermal image types required training a neural network to learn
when the features from the two types of images were highly correlated, making it possible for a matching system trained with visible imagery to be applied to thermal imagery.
The approach will allow the military to expand its facial recognition capabilities without developing custom software.
Read the full paper at is.gd/GCN_ARL.•
BY TROY K. SCHNEIDER
Jose Arrieta is in love with the data layer. In his role as the Department of Health and Human Services’ associate deputy assistant secretary for acquisition, Arrieta oversees the acquisition function across the highly decentralized agency. Like many government systems,
HHS’ are often siloed and expensive to integrate, so the business processes
are rife with manual tasks. And because the time and money required to modernize the systems wholesale are prohibitive, the inefficiencies persist.
Arrieta, however,
believes that a robust,
well-structured and
immutable data layer offers a way
out and could allow HHS and other agencies to automate and improve their business processes while running on top of legacy systems. Specifically, he said blockchain can power that middleware for a wide range of government operations.
“I believe that there’s a unique opportunity now to actually layer your existing systems with high-powered middleware,” Arrieta said at an
April FCW event on automation and cybersecurity. “You can let your existing operating environment deliver services and you can build microservices off
of your data layer to modernize your business systems at the same time.”
In his previous job at the General Services Administration, Arrieta piloted such a solution for the IT Schedule
70 FASt Lane process, and “we were conservatively looking at lowering our
opex costs by 80 to 90 percent,” he said. He sees similar opportunities at HHS and elsewhere.
The key, he said, is to have a
data layer that comes with “CIA”
— confidentiality, integrity and accessibility. A blockchain distributed ledger checks all three boxes.
Agencies can “build microservices directly off of that data
10 GCN APRIL/MAY 2018 • GCN.COM
Jose Arrieta, Department of Health and Human Services
layer to drive business outcomes,” he said. “And you can test those microservices at a very low cost...all while you’re leaving your existing environment in place.” That can allow agencies to modernize processes now, “and then you can slowly
move off your existing
IT architecture” as the resources to do so become available.
Deciding what to include and extracting it from existing systems
are significant challenges. He recommended starting small, with
a specific process “that you actually own,” and cited basic access controls as an example.
KPMG Principal Tony Hubbard, who also spoke at the event, agreed that access control was “a perfect example where automation can really lend
itself to support an effort.” He also urged agencies to look at their Risk Management Framework efforts and the challenges of “taking systems to the accreditation stage.”
“It’s very cumbersome, with a lot
of manual processes,” Hubbard said. “For those of us who work in the cybersecurity field, there’s probably no better example than RMF that can be automated.” •
PHOTO COURTESY OF ERIC PROCTOR, WILLIAM PARKS AND BENJAMIN S. RIGGAN