Indonesia’s Akurat Satu rises in latest NIST facial recognition accuracy rankings
Indonesia biometrics developer Akurat Satu has finished in the top 25 for facial recognition accuracy with wild images in testing by the U.S. National Institute of Standards and Technology (NIST).
Akurat Satu submitted its algorithm to NIST’s Face Recognition Vendor Test (FRVT) for the first time for the August report, and in the September update finished with the 25th-best accuracy for matching Wild Faces at a false non-match rate (FNMR) of 0.00001. The company also finished 26th in matching Border photos, 33rd in Visa photos, 37th in the Visa Border category, 40th in mugshots after 12 or more years and 42nd in more recent mugshots.
“Our biometrics system and technology have been around since 2011 and are used by various state institutions, law enforcement, and companies,” Akurat Satu’s Director Christian Kurniawan stated, according to the Jakarta Globe.
The article notes the importance of wild image accuracy for applications like border control, airport security, law enforcement forensics, digital KYC and public surveillance systems.
Clients of Akurat Satu include the National Anti-Narcotics Agency (BNN), the National Police (Polri), and the National Counter-Terrorism Agency (BNPT). The company’s technology is also used by partners Asli RI and VeriJelas.