In the latest NIST Evaluation of Latent Fingerprint Technologies (ELFT), ROC achieved industry-best Rank-5 forensic accuracy and the fastest search speeds of any vendor tested, confirming its growing influence in real-world forensic scenarios where speed and accuracy matter most. With matches completed in just 386 seconds—roughly 20x faster than the average of its Western peers—ROC gives law enforcement, defense, and forensic agencies the power to search larger databases, surface better leads, and close cases faster with trusted American-made biometrics.

Why Latent Fingerprints Are So Challenging

ROC, the only American-made multimodal biometrics and Vision AI provider, today announced standout results in the latest NIST Evaluation of Latent Fingerprint Technologies (ELFT). The results confirm ROC’s position as the fastest vendor in the benchmark while delivering forensic-grade precision in the most operationally relevant test scenarios. NIST ELFT is the world’s most rigorous benchmark for assessing both the accuracy and efficiency of automated latent fingerprint recognition algorithms.

Latent prints are often partial, low-quality impressions lifted from crime scenes. Because of their incomplete and degraded nature, they are among the most difficult forms of biometric evidence to process—yet also some of the most valuable, since they often provide the only physical link between a suspect and a crime scene. For investigators, a successful latent match can mean the difference between stalled leads and actionable intelligence, making advances in accuracy and speed critical for solving cases.

Forensic Accuracy in Action: FBI and DoD Results Explained

In the FBI’s Solved Dataset #1, which consists of 285 operational probes collected from real crime scenes and annotated with EFS examiner markups¹, ROC achieved the lowest Rank-5 search error rate of any vendor tested, confirming its leadership in examiner-markup conditions that simulate authentic casework. Rank-5 accuracy measures how often the correct identity appears within the top five candidates returned by the system, reflecting the way forensic examiners work in practice and making it one of the most realistic measures of operational performance.

Rank-5 error rate on the FBI Solved Dataset with Extended Feature Set Markup for all vendors in the ELFT benchmark, along with search speeds in seconds.

ROC also delivered the second-lowest Rank-5 error rate of any Western provider in the DoD Dataset #1, a benchmark designed to evaluate performance at scale in defense-level search conditions. The DoD Dataset is also significant because it represents a large-scale, real-world operational test case: the fingerprints were not laboratory-prepared under ideal conditions, but collected during actual DoD operations. The result is a direct measure of how AFIS systems perform in hazardous field conditions.

Rank-5 error rate on the Department of Defense Dataset for all Western vendors in the ELFT benchmark, along with search speeds in seconds.

“This is just the beginning of what our next-generation latent fingerprint technology can do. With each improvement, law enforcement and intelligence missions move closer to tools that truly serve their needs. Investigators can now search larger databases faster, widening search parameters and ensuring no potential lead is left behind… All powered by American-built technology that breaks the multi-decade, foreign grip on U.S. government biometric systems.”

 

– Dr. Brendan Klare, Chief Scientist and Co-Founder, ROC

Business as Usual: ROC Speed and Accuracy Gains

At the same time, ROC posted the fastest query times in the benchmark—386 seconds across an estimated 30 million fingerprint gallery—compared to an average of 8,029 seconds for other Western vendors. This 20x speed advantage enables agencies to search exponentially larger databases without sacrificing accuracy.

ROC’s newest latent fingerprint algorithm also delivered major year-over-year improvements, with error rates reduced by nearly 2x overall and by as much as 10x on certain ELFT benchmarks.

“At ROC, we’re proving that next-generation fingerprint capabilities are achievable by combining the latest advances in Vision AI with traditional biometrics research. In practice, this has already delivered search speeds that are orders of magnitude faster, coupled with best-in-class accuracy. As we continue to draw on our expertise across both AI/ML and Biometrics, we expect our next series of algorithm improvements to further solidify ROC’s role at the forefront of fingerprint recognition solutions.”

 

– Dr. Joshua Engelsma, Principal Scientist, ROC

An Operational Edge in Latent Fingerprint Recognition

Forensic examiners and investigators rely on latent fingerprint systems not only for accuracy, but for speed. By delivering both simultaneously, ROC provides agencies with a decisive operational advantage:

• Defense and intelligence missions can search expansive databases in near real time.
• Law enforcement agencies gain faster investigative leads to solve high-sensitivity crimes.
• Civil identity programs and national ABIS deployments benefit from ROC’s proven ability to scale accuracy and speed in large databases.
• Mission-critical environments gain from ROC’s balance of accuracy, scalability, and American-built trust.

Unlocking a Stagnant Fingerprint Market

Only two years in the making, ROC’s highly efficient fingerprint algorithm is already demonstrating the possibilities of a next-gen approach: legacy hardware limitations that once forced agencies to search only subsets of fingerprints are simply no longer relevant.

As ROC’s accuracy continues to improve, new levels of performance are emerging in investigative workflows, giving latent examiners a leg up by expanding search capacity, responsiveness, and ultimately, operational outcomes. 

Looking ahead, ROC expects continued and consistent progress. Planned improvements in score normalization and automated watchlist matching (DET results) will further strengthen accuracy, with another latent fingerprint algorithm release anticipated before the year’s end.

Learn more about ROC fingerprint recognition

Understanding NIST ELFT

The NIST Latent Fingerprint Evaluation (ELFT) is the authoritative one-to-many benchmark run by the National Institute of Standards and Technology. It measures how effectively automated latent fingerprint systems perform, using challenging, partial, and unmarked images as probes against large reference databases. Importantly, ELFT compares “image only” searches with systems that also use examiner-marked features, helping agencies understand the added value of expert markups in improving accuracy. The program is widely regarded as the most trusted and rigorous benchmark for latent fingerprint matching, providing transparent performance evaluation across both accuracy and speed. Learn more at www.nist.gov

¹See “Benefit of reviewing 5 ranks” placing for ROC in the nutrition report.