EyeDetect Alternatives:
6 Options Ranked

Six technologies compete with Converus EyeDetect for credibility assessment: FACS-based facial Action Unit analysis, polygraph, thermal imaging, P300 brain fingerprinting, fMRI, and voice stress analysis. EyeDetect itself measures one channel — ocular-motor response — during a structured 30-minute computer test, with results processed on Converus servers in the US. The ranking below is for one specific requirement: continuous, deployable credibility assessment in security-cleared environments. That requirement rewards non-contact operation, on-premise processing and the absence of a structured test protocol. A different requirement produces a different order, and the reorderings are set out at the end. No technology here detects lies; each infers deception from an indirect physiological correlate, and all of them degrade sharply when applied to low-base-rate population screening rather than specific-incident testing, which is the single error this page exists to prevent.

The ranking at a glance

1. FACS-based facial Action Unit analysis

FACS-based analysis measures 44 facial Action Units per frame, including micro-expressions and suppression patterns, from a standard RGB camera at 720p. Against EyeDetect specifically, the differences that matter are channel count (44 against one), the absence of any structured test protocol, and processing location — EchoDepth runs entirely on-premise and is air-gap deployable, while EyeDetect results are processed on Converus servers in the US. See the full EchoDepth vs EyeDetect head-to-head.

Its honest limitation: the face must be visible and adequately lit, and while FACS as a measurement framework has forty years of peer-reviewed validation, the literature tying the wider Action Unit set specifically to deception is correlational, exactly as it is for every other option here.

2. Polygraph

Polygraph remains the institutional default and has an accreditation and legal history no other option can match. That is a real procurement advantage, and it is why polygraph ranks second here despite its evidence problems. Those problems are substantial: the US National Academy of Sciences concluded in 2003 that accuracy is insufficient for security screening, with false negative rates of 20–47% in that context. See EchoDepth vs Polygraph.

3. Thermal imaging

Thermal imaging measures periorbital blood flow as an arousal proxy, with roughly 75–80% laboratory accuracy reported by Pavlidis and colleagues in Nature in 2002. It is the only option here that requires no subject cooperation at all, which makes it attractive for border and checkpoint screening — but specificity is low and performance degrades with ambient temperature, recent exertion and clothing, all uncontrolled in exactly those settings.

4. P300 / brain fingerprinting

P300 methods detect whether a subject recognises a probe stimulus, which is a different and in some ways more tractable question than whether they are lying. Where crime-relevant detail is genuinely held back from the public, this is a real capability EyeDetect lacks. In practice it requires an EEG cap, a carefully constructed stimulus set and a specialist operator, and independent replication of the accuracy claims is minimal.

5. fMRI

fMRI reports the highest laboratory accuracy of any technology discussed here, around 90%. It has never been validated in field conditions, requires the subject to lie immobile in a scanner costing millions, and was excluded as evidence in US v. Semrau (2010). It is a research instrument and should be treated as one.

6. Voice stress analysis

Voice stress analysis is listed for completeness rather than as a recommendation. Independent assessments including a 2002 US Department of Defense Polygraph Institute study and subsequent National Institute of Justice field studies found it performs at chance level. It is still actively marketed, which is the only reason it appears on this list.

When EyeDetect is still the right choice

Three situations where EyeDetect beats everything above, including EchoDepth:

• You need a short, standardised test across a defined population. EyeDetect takes around 30 minutes, runs the same protocol for every subject, and needs very little examiner time. Polygraph cannot match that throughput, and continuous behavioural monitoring answers a different question entirely.
• Examiner availability is your binding constraint. EyeDetect was designed specifically to reduce examiner dependency, and it does. If you have more subjects than trained examiners, that is a genuine operational advantage.
• Cloud processing is acceptable and you want an established commercial product. If your environment permits US-hosted processing, EyeDetect is a mature, supported, commercially deployed system with a real installed base — which is not nothing when compared against research instruments.

Conversely, EyeDetect is ruled out where processing must remain on-premise or inside a SCIF, where assessment must occur during natural interaction rather than a scheduled test, or where continuous rather than point-in-time assessment is the requirement.

How the ranking changes with the requirement

The order above is specific to continuous, deployable assessment in cleared environments. Change the requirement and it reorders:

• Maximum legal familiarity → polygraph first, then EyeDetect, then FACS-based analysis.
• Highest throughput at a physical checkpoint → thermal imaging first, then FACS-based analysis, then everything else.
• Testing recognition of specific withheld detail → P300 first; nothing else addresses this directly.
• Laboratory accuracy alone, ignoring deployability → fMRI first.
• Lowest cost per subject with no accuracy requirement → voice stress analysis, which is precisely why it persists.

The broader point for procurement: no technology on this list resolves the low-base-rate problem that makes population screening unreliable. Comparing them on headline accuracy figures drawn from specific-incident studies, and then deploying the winner for screening, is the single most common error in this category. See polygraph alternatives compared for the full evidence review.