Uniting Citizens with Government Databases: OneRegistry in Sri Lanka
Title: iBeta's Expansion into Unbiased Biometric Testing
Kickin' it old-school: iBeta Quality Assurance, the venerable vet in the realm of quality checks, has just upped the ante. They're diving headfirst into evaluating biometric systems for demographic bias – and here's the lowdown on their latest service offering.
The low-down: iBeta's new service is all about determining whether biometric systems perform smooothly across a spectrum of demographic groups, backed by the ISO/IEC 19795-10 standard. This bad boy got finalized and unveiled last year, focusing on guess what? That's right – quantifying the performance variations across different demographics, my friends![1]
Breaking it down: The testing involves parsing biometric data into categories like age, gender, and skin tone (yep, using the Monk scale). After leveling the playing field by adjusting for the number of subjects per category, iBeta compares the performance rulers across groups, looking for evidence of potential bias.
Addressing the elephant in the room: Algorithmic bias has been a hot topic lately, as biometrics make their way into both the public and private sectors like a wireless skin tag. With this service, providers can show that their biometric solutions play fair and treat all demographics equally[1].
Certification and cred: To top it off, iBeta snatched accreditation for this testing under NIST’s National Voluntary Laboratory Accreditation Program (NVLAP). In other words, they've earned a badge of honor as a trailblazer in biometric testing[1].
[1] iBeta Quality Assurance expands biometrics testing to address demographic bias, finding bias in face recognition for people with darker skin, (2021, January 14). Retrieved March 15, 2021, from https://www.ibeta.org/ibeta-quality-assurance-expands-biometrics-testing-to-address-demographic-bias/#:~:text=iBeta%20uses%20the%20Monk%20scale,for%20peoples%20with%20darker%20skin.&text=iBeta's%20service%20aims%20to%20help,biometric%20technologies%20as%20they%20become%20more%20widespread.
- iBeta's venture into the unbiased biometric testing domain utilizes data-and-cloud-computing technologies to analyze biometric data across various demographic groups, such as age, gender, and skin tone, ensuring fairness in performance.
- By employing advanced technology and adhering to the ISO/IEC 19795-10 standard, the biometric testing service offered by iBeta, recently accredited under NIST’s National Voluntary Laboratory Accreditation Program (NVLAP), aims to mitigate potential data bias in biometric technologies as they become more widespread in both public and private sectors.
