ROC highlights facial recognition accuracy, efficiency scores as school expansion planned
A series of updates from Rank One Computing highlight the accuracy of the company’s facial recognition software, and its commercial prospects in the school safety market.
A pilot of ROC’s facial recognition at West Fairmont Middle School in West Virginia is continuing, with officials from the Department of Homeland Security, local and state lawmakers, and superintendents from other school districts visiting to see the technology in action, The Dominion Post reports (via Yahoo).
The district is building up its database of parents, vendors, and other visitors to the school.
Officials are now working towards the implementation of ROC’s facial recognition in the county’s 19 other schools, according to the report.
NIST FRVT success
The latest version of the ROC SDK showed major improvements across the board in the 1:N benchmark test by the National Institute of Standards and Technology dated February 2, 2023.
The company says it scored the best combination of accuracy and efficiency of any global technology provider, and the lowest error rate among Western developers. The ROC algorithms error rate was cut in half in just five months, according to a company announcement.
“We are pleased but not surprised to see that our latest edition of ROC SDK leads the pack on facial recognition,” says CEO Scott Swann. “Through continuous innovation, ROC.ai will only continue to shape the future of the global biometrics landscape with objectively superior computer-vision capabilities.”
A company blog post announcement compares the performance of the ROC SDK v2.4 favorably to other leading competitors in the visa border category.
ROC says that while Chinese companies are leading seven out of eight categories, its own accuracy is often within 0.1 percent of the global leader. In efficiency, as expressed by template size, comparison time, template generation time and amount of memory needed to run the algorithm, ROC ranks well ahead of its closest market competitors by error rate.
The post also delves into ROC’s normalized error rates across different demographics, and the improvement of its accuracy over time.
New COO appointed
Tony Brown has been named Chief Operating Officer of ROC AI following several years of working with the company as a contractor and consultant.
Brown is the co-founder and former COO of consultancy Acute Insight. He brings experience in biometrics, identity management, and science and technology portfolio management, according to the announcement.