[ad_1]
Parabon’s technology “can’t tell you the exact millimeters between the eyes or the ratio between the eyes, nose and mouth,” Greytak said.Without that level of precision, facial recognition algorithms can’t provide accurate results — but getting such precise measurements from DNA requires fundamentally new scientific discoveries, she said, and “trying to predict at that level.” The paper didn’t have a lot of success.” Greytak says Parabon can only predict the approximate shape of a face (although The scientific feasibility of such predictions has also been questioned).
Police have been known to draw forensic sketches based on witness descriptions through facial recognition systems.One Georgetown Center for Legal Privacy and Technology 2019 Research It was found that at least six police agencies in the US “allow, if not encourage” the use of forensic sketches, either hand-drawn or computer-generated, as input photos for facial recognition systems.AI experts warn that such a process is likely resulting in lower accuracy.
Corsight has also been criticized in the past for exaggerating the capabilities and accuracy of its facial recognition system, which it calls “the most ethical facial recognition system in extremely challenging conditions,” according to the slideshow Presentations available online. in a IPVM Technology Demonstration In November, Corsight CEO Watts said that Corsight’s facial recognition system can “identify people who are wearing masks — not just masks, but ski masks.” The IPVM reports that using Corsight’s face recognition system on masked faces The AI got a 65% confidence score, which Corsight itself measures how likely a captured face is to match in its database, noting that the mask is more accurately described as a balaclava or neck gaiter, rather than just a mouth and eye cutout ski mask.
There are broader problems with the accuracy of facial recognition technology good–record on file (including through MIT Technology Review).They are more noticeable when the photo is poorly lit or taken at extreme angles, and When the subject has dark skin, is female, or is very old or very young. Privacy advocates and the public have also criticized facial recognition technology, especially as Clearview AI Crawl social media as part of the matching engine.
The use of the technology by law enforcement is of particular concern — Boston, Minneapolis and San Francisco are among the many cities that have banned its use. Amazon and Microsoft have stopped selling facial recognition products to police organizations, and IBM has taken its facial recognition software off the shelves.
“fake-science”
“The idea that you’re going to be able to create something with the granularity and fidelity you need to run a face-matching search — to me, that’s ridiculous,” said Albert Fox Kahn, a civil rights attorney and executive director of the Surveillance Technology Oversight Project , who has worked extensively on issues related to facial recognition systems. “That’s pseudoscience.”
Dzemila Sero, Researcher Computational Imaging Group The head of Centrum Wiskunde & Informatica, the Netherlands’ national institute for mathematics and computer science, says the science behind such a system is not fully developed, at least not yet publicly. Citing a 2017 study by Human Longevity, Sero said the catalog of genes needed to accurately delineate a face from a DNA sample is currently incomplete.
In addition, factors such as environment and aging have a major impact on faces that cannot be captured by DNA phenotyping, and studies have shown that individual genes have less of an impact on facial appearance than gender and ancestry. “Premature attempts to implement this technology risk undermining trust and support for genomic research and garnering no societal benefit,” she told MIT Technology Review in an email.
[ad_2]
Source link
