Neurotechnology tops NIST iris recognition evaluation

By AI, Created 3:11 PM UTC, May 27, 2026, /AGP/ – Neurotechnology said its latest iris recognition algorithm ranked first across all reported accuracy metrics in NIST’s IREX 10 evaluation, which tests one-to-many identification performance on large-scale datasets. The company said the result strengthens its position in biometric benchmarking and will feed into future MegaMatcher license upgrades.

Why it matters: - Neurotechnology’s result puts its iris recognition algorithm at the top of a widely watched benchmark used to gauge performance in real-world identification and investigation systems. - The ranking matters for government and security applications that depend on fast, accurate biometric matching, including national ID, border control and forensic review. - Neurotechnology said the latest submission also tripled matching speed versus the nearest competitor, pairing accuracy with throughput.

What happened: - Neurotechnology said its latest iris recognition algorithm, submission neurotechnology_020, ranked first across all reported accuracy metrics in NIST’s Ongoing Evaluation of Iris Recognition 10, or IREX 10. - NIST IREX 10 measures iris recognition performance in one-to-many identification and investigation scenarios using large-scale, real-world datasets. - The evaluation used 1 million iris images from 500,000 individuals. - Neurotechnology said the results covered both single-eye and two-eye accuracy testing.

The details: - The evaluation used two main metrics: Detection Error Tradeoff, or DET, Accuracy and Ranked Accuracy. - DET Accuracy measures performance in automated identification systems such as national ID programs and border control. - Neurotechnology achieved the best results at the headline FNIR@FPIR 0.01 operating point and across the reported DET curves for single-eye and two-eye identification. - Ranked Accuracy measures how often the correct identity appears among top search results in human-in-the-loop investigations. - Neurotechnology’s algorithm posted the lowest miss rates at Rank 1, Rank 10 and Rank 100 in both single-eye and two-eye scenarios. - Neurotechnology said the submission was the most accurate among all participating developers across all four evaluations.

Between the lines: - The result reinforces a broader industry shift toward biometric systems that are judged on both precision and speed, not accuracy alone. - Neurotechnology said the outcome shows the next generation of iris recognition must combine peak precision with high-speed performance. - The company is also active in other NIST evaluations, including MINEX III, PFT III, FRIF TE E1N, FRTE, FATE, ELFT and SlapSeg III, where it says its algorithms often rank among the top performers. - That broader testing footprint suggests Neurotechnology is using NIST benchmarks as a key proof point across multiple biometric categories.

What’s next: - Neurotechnology said the new algorithm will be included in the next MegaMatcher license upgrades. - MegaMatcher products are used in national identity and voter management, law enforcement and border security. - The company’s public-facing channels include LinkedIn, Facebook, YouTube and X.