IREX IX Results and Previous IREX Evaluations
In 2018 Neurotechnology's iris recognition algorithm has been judged by the National Institute of Standards and Technology (NIST) as the second most accurate among the participants. The accelerated version of the algorithm was nearly 50 times faster than any other matcher in the NIST IREX IX evaluation. Our comments on IREX IX participation contain more details about the results.
MINEX III Compliance and previous MINEX evaluations
In 2018 Neurotechnology's fingerprint template generator algorithm has been ranked the first in the NIST MINEX interoperability category; the fingerprint matching algorithm ranked second and, when combined, the two become the most accurate high speed fingerprint recognition system. Also the latest fingerprint algorithm for smart cards submission has demonstrated significant improvement in reliability with proven outstanding template generator at enhanced performance. In 2017 MegaMatcher SDK fingerprint technology was ranked as the first most interoperable matcher and the fourth most accurate native template matcher vendor among all MINEX III compliant matchers. In 2016 MegaMatcher on Card SDK fingerprint matching algorithm for smart cards also successfully passed MINEX III evaluation. Our comments on MINEX III participation contain more details about the results.
In 2014 MegaMatcher SDK fingerprint technology was recognized by the NIST as fully MINEX compliant and placed second in the Ongoing MINEX ranking for fingerprint matching algorithms.
In 2007 previous version of MegaMatcher SDK was one of the several algorithms worldwide recognized as fully MINEX compliant for both fingerprint template encoding and matching.
In 2018 Neurotechnology has been ranked among 8 most accurate face recognition algorithm vendors out of 39, with tenth most accurate algorithm out of 78 in the FRVT leaderboard. The submission was also ranked as one of the best in two difficult scenarios, with second most accurate result on a complex dataset collected from operational photos related to ongoing criminal investigations, and fourth most accurate result with unconstrained, photojournalism-style photos. Our comments on FRVT participation contain more details about the results.
Kaggle Competition on DNN-based species classification
In 2017 Neurotechnology researchers won first place in a Kaggle competition with deep neural network based computer vision solution for classifying fish species.
Read the press release for more information.
PFT II (Proprietary Fingerprint Template) Evaluation
In 2017 Neurotechnology fingerprint algorithm was submitted to NIST Proprietary Fingerprint Template Evaluation II. The algorithm's template matching accuracy was among the best participants in most of the experiments. See our comment for more information.
FpVTE 2012 and FpVTE 2003 (the Fingerprint Vendor Technology Evaluations)
In 2015 Neurotechnology's fingerprint identification algorithms have been judged by the National Institute of Standards and Technology (NIST) as one of the fastest and most accurate among the participants. Our comments on FpVTE 2012 participation contain details about the results in each category.
Previously, Neurotechnology participated in FpVTE 2003 under the name Neurotechnologija and showed one of the best reliability results in the Middle Scale Test. See the FpVTE 2003 web site for a detailed report of the evaluation results.
FIVE (Face in Video Evaluation)
In 2015 Neurotechnology face recognition engine was submitted to the NIST Face in Video Evaluation (FIVE). In average the submitted algorithm was ranked among top 8 most accurate face recognition algorithms out of 16 vendors. See our comment for more information.
WSQ 3.1 Certification
In 2011 FBI certified Neurotechnology's implementation of WSQ image format support. Certificates and additional information are available.
FVC2006, FVC2004, FVC2002 and FVC2000 results
Neurotechnology participated in the Fingerprint Verification Competition several times and won numerous medals for reliability and performance. See the FVC2006 participation results, as well as FVC2004, FVC2002 and FVC2000 results for more information.
Comments on competition results
The FpVTE protocol was strict and did not allow use of some of our advanced algorithm features, which, in a real world application, would further increase the recognition quality. Particularly, the MST set contained images from different scanners, but each particular image scanner model was not disclosed. In a real world scenario, specific parameters would be set for each specific scanner type. This would allow the algorithm to perform at an even higher accuracy level.
Another such real world example that was not simulated in the FpVTE protocol is the ability to generate globalized or generalized features templates by capturing several images from the same finger and combining the templates into a single features set. Using a generalized features set can significantly improve the algorithm's reliability and produces improved matching scores. In the FpVTE MST set such a method could not be used, as only two matched fingerprints were allowed for consideration.
The FVC protocol is very useful for comparing different vendors' algorithms, however it only allows comparison of verification (1-to-1 matching) but not identification (1-to-many matching) results. One of the strongest capabilities of Neurotechnology's algorithms is fast, reliable identification, therefore a 1-to-many test would better reflect our real algorithm ranking among the participants.
FVC uses databases that are not from real applications (more information), but rather uses fingerprint sets which had been specially collected for the competition (some with certain distortion or noises highlighted). In this way, various distortion and noise statistics of the fingerprints did not correspond to real world application statistics, and vendors' results may be not completely adequate to apply to real life situations.
Like the FpVTE, the FVC did not allow us to generate globalized or generalized features templates by capturing several images from the same finger and combining the templates into a single features set. Using a generalized feature set can significantly improve the algorithm's reliability and produces improved matching scores. In the FVC such a method could not be used, as information from only two matched fingerprints was allowed for consideration.