In short
Ayonix algorithms have been submitted to NIST face recognition evaluations in both the 1:1 verification and 1:N identification tracks. NIST publishes comparative results and does not certify or endorse vendors — no NIST certification exists. This page links NIST's own records rather than summarising them.
What the NIST face recognition evaluation is
The National Institute of Standards and Technology, a United States federal agency, runs an ongoing independent evaluation of face recognition algorithms. Developers submit algorithms as compiled software; NIST runs them against large sequestered datasets the developers never see, and publishes comparative reports.
It matters because almost nothing else in this market is independent. A vendor-published accuracy figure is measured by the vendor, on data the vendor selected, at a threshold the vendor chose. NIST removes all three of those degrees of freedom at once:
- Algorithms are tested on data the submitter has never seen, so tuning to the test set is not possible.
- All submitted algorithms face the same data, which makes cross-vendor comparison meaningful.
- Results are published openly, including error rates broken down by demographic group.
- Submissions are dated, so a result is tied to a specific algorithm version at a specific time.
What Ayonix’s participation means, precisely
Ayonix algorithms appear in NIST face recognition evaluation publications across multiple test cycles, in both the 1:1 verification and 1:N identification tracks. That is the complete claim, and this page will not extend it.
NIST publishes comparative evaluations and does not endorse vendors. There is no such thing as a NIST certification, a NIST approval or a NIST licence for face recognition. Any vendor presenting a badge implying otherwise is describing something the agency does not issue.
The record
Where to check this yourself
NIST publishes its participant lists and per-algorithm report cards openly. Any vendor’s participation claim, this one included, should be checked at NIST rather than taken from the vendor’s own site.
| Evaluation | Algorithm | Submitted | What the NIST record shows |
|---|---|---|---|
| FRTE / FRVT 1:1 verification | ayonix_000 | 22 June 2017 | Developer, algorithm type and submission date, plus the full report card NIST generated for it. |
| FRTE / FRVT 1:N identification | ayonix_0, ayonix_1, ayonix_2 | Traces from 2018 | Three identification algorithms, each with its own published report card linked from the results page. |
| FATE Age Estimation & Verification | ayonix_001 | 23 July 2026 | Listed by NIST as “Ayonix (JP)”, with a published per-algorithm report card. No error figure from it is quoted here; the NIST results page is linked below as the source. |
NIST FRTE 1:1 verification — participant list
Ayonix is listed there as Ayonix (JP).
NIST FRTE 1:N identification — results and report cards
Links the per-algorithm report cards, including those for ayonix_0, ayonix_1 and ayonix_2.
NIST FATE Age Estimation & Verification — results
Lists ayonix_001 as “Ayonix (JP)” with its own report card. This is the source for the FATE row above.
NIST Face Technology Evaluations programme
The programme itself, for the current FRTE and FATE tracks and the reports published under them.
NIST FRTE 1:1 verification report (PDF, 88 MB)
The published report itself, which is the appropriate independent source for any comparative figure.
No figure from these reports is quoted on this page — not the error rates, and not the position in the ranked tables. The reports are linked in full instead. A vendor summarising its own evaluation results is choosing which results to summarise, and a reader is better served by the source than by the summary.
How to read one
Six things to check in any NIST report
Useful whether you are reading the Ayonix records or any other vendor’s — and the reason a vendor’s one-line summary of its own result is rarely enough.
Identify the track
1:1 verification and 1:N identification are different problems with different reports. A strong result in one says very little about the other.
Check the submission date
Reports accumulate over years, and a vendor’s current algorithm may not be the one in the row being quoted at you.
Read both error rates together
An algorithm can be tuned to look excellent at one by being poor at the other. A figure quoted alone conceals the trade it was bought with.
Find the demographic differentials
Error rates vary across age, sex and skin tone. A single headline figure averages that away entirely, and it is usually the number that matters most.
Note the dataset
Visa photographs, mugshots and border-crossing images produce very different numbers for the same algorithm.
Treat it as evidence, not prediction
It describes the algorithm on that data. Your camera placement, lighting and population are not in the report, and they are what decide your outcome.
Terminology
FRVT, FRTE, and the four error rates
The same programme under two names, and four measurements a vendor can choose between. Knowing which is which is most of what it takes to read a claim critically.
FRTE 1:1
Verification
Are these two images the same person? Reported with FMR — the false match rate, how often two different people are called the same — and FNMR, the false non-match rate, how often the same person is not recognised. Neither moves with gallery size.
FRTE 1:N
Identification
Which of the N enrolled people is this, if any? Reported with FPIR and FNIR, the false positive and false negative identification rates. Unlike FMR and FNMR these move with the size of the gallery being searched, which is why an identification result cannot be inferred from a verification one.
FATE
Face Analysis Technology Evaluation
Quality, morph detection, age estimation and similar tasks that are not recognition, evaluated separately. Worth knowing about so that a result from it is not mistaken for a recognition result.
FRVT and FRTE are the same ongoing programme under different names. NIST ran its face recognition evaluations as the Face Recognition Vendor Test for many years and has since reorganised the tracks as the Face Recognition Technology Evaluation. Older reports and URLs still carry the FRVT name, which is why a search turns up both — and why a vendor citing FRVT is not necessarily citing something retired.
The registered claims
What this page is permitted to state
Each of these is registered with its source, its verification date and a named reviewer before it may be published. Each also states what it does not establish.
Ayonix algorithms have been submitted to NIST face recognition evaluations
- Source type
- Government or standards body
- Verified
- 2026-09-11 · Dr Sadi Vural, Chief Executive Officer
What this does not establish
NIST publishes comparative evaluations and does not endorse, certify, approve or accredit vendors. Participation means an algorithm was measured on data the submitter did not choose. It is not a statement about performance at any particular site.
One verification algorithm, ayonix_000, submitted to the NIST 1:1 verification track on 22 June 2017
- Source type
- Government or standards body
- Verified
- 2026-09-11 · Dr Sadi Vural, Chief Executive Officer
What this does not establish
No error rate or ranking from the report is quoted. A 1:1 result does not predict 1:N behaviour.
Three identification algorithms, ayonix_0, ayonix_1 and ayonix_2, with traces in the NIST 1:N identification track from 2018
- Source type
- Government or standards body
- Verified
- 2026-09-11 · Dr Sadi Vural, Chief Executive Officer
Ayonix did not participate in the standalone NIST FRVT demographic effects study
- Source type
- Ayonix first-party statement
- Verified
- 2026-09-11 · Dr Sadi Vural, Chief Executive Officer
What this does not establish
Demographic figures inside the ayonix_000 report card are part of the standard NIST template and are not a demographic-study result.
Frequently asked questions
Is Ayonix NIST certified?
No, and neither is any other company. NIST runs face recognition evaluations and publishes comparative results, but it does not certify, approve, accredit or endorse vendors — no such certification exists. What Ayonix can accurately say is that its algorithms have been submitted to NIST face recognition evaluations in both the 1:1 verification and 1:N identification tracks and appear in the published evaluation materials. Any vendor presenting a NIST certification badge is describing something the agency does not issue.
Which NIST evaluations has Ayonix participated in?
One verification algorithm, ayonix_000, submitted to the 1:1 track on 22 June 2017; three identification algorithms — ayonix_0, ayonix_1 and ayonix_2 — with traces in the 1:N track from 2018; and ayonix_001 in the FATE Age Estimation & Verification track, submitted 23 July 2026. The programme has run under the names FRVT and, more recently, FRTE and FATE. NIST publishes its own participant lists and per-algorithm report cards, and those are the appropriate place to verify this rather than taking it from a vendor page.
What accuracy did Ayonix achieve in NIST testing?
This site publishes no accuracy figure, in NIST terms or any other. An error rate without its dataset, operating threshold and demographic breakdown cannot be reproduced, and quoting a favourable slice of a NIST report is the most common way that context gets lost. The reports are linked in full instead, because a reader is better served by the source than by a vendor’s summary of its own results.
What is the difference between FRVT and FRTE?
They are the same ongoing NIST programme under two names. Face Recognition Vendor Test was the long-running designation; NIST later reorganised the work as Face Recognition Technology Evaluation, with Face Analysis Technology Evaluation covering attribute tasks that are not recognition. Reports and URLs exist under both names, which is why a search for any vendor should cover each.
Does a good NIST result mean the system will work at our site?
Not directly. A NIST result measures an algorithm on sequestered data under controlled conditions, which is strong evidence that the vendor submitted to independent scrutiny. It does not predict performance with your cameras, your lighting, your mounting angles and your population. Both statements are true at once, and vendors tend to quote only the first. Run a pilot on your own site before committing, whatever the published results show.
Why do some vendors advertise a NIST ranking?
Because NIST reports contain ordered result tables, and a vendor can truthfully report placing highly in one table on one date. It becomes misleading when the track, dataset, submission date and demographic breakdown are omitted — a different table from the same programme often tells a different story about the same algorithm. That is why no ranking appears on this site.
Did Ayonix take part in the NIST demographic effects study?
No. Ayonix does not appear in that standalone study. Demographic figures do appear inside the ayonix_000 verification report card, but they are part of NIST’s standard report template rather than a result from the demographic study — which is a different thing, and one worth stating explicitly because the distinction is easy to blur and commonly is.
Next step
The number that describes your site
A NIST report describes an algorithm on sequestered data. A pilot on your own cameras, with your own population and acceptance criteria agreed in advance, describes your deployment. Only the second is a number you can plan against.