“The platform allows users to upload photographs from the Civil War, and then cross-references them using facial recognition technology, presenting possible matches for end users who might then be able to piece together the subjects’ identities.”
A joint project between military historians and Virginia Tech engineers is using facial recognition technology to identify soldiers from the Civil War.
As Slate reports, the project was spearheaded by Kurt Luther, a computer scientist who researches the Civil War as a hobby. Collaborating with Ron Coddington, the editor of Military Images magazine; Paul Quigley, the director of the Virginia Center for Civil War Studies, and student researchers from Virginia Tech, he launched a web app called Civil War Photo Sleuth this summer. The platform allows users to upload photographs from the Civil War, and then cross-references them using facial recognition technology, presenting possible matches for end users who might then be able to piece together the subjects’ identities.
While contemporary facial recognition technology isn’t always well-suited to archival photographs, and especially the many bearded visages presented in photos of soldiers from the Civil War, Coggington has estimated that roughly 40 million photos were taken during the calamitous conflict, so even if only a portion of those survive today, there’s still a wealth of raw material to work with. And it looks like it’s already proving effective, with the CWPS having logged 88 reported identifications in August, its first month of operation.
It’s a good example of an alternative application of contemporary facial recognition technology beyond the surveillance and mobile authentication use cases that are so popular today, one that is uncontroversial and aimed at the broader public good. A century and a half after America’s most painful conflict, there’s still more to learn, and biometric technology can help.
November 16, 2018 – by Alex Perala