> Skip to main content
Credibility Assessment

EchoDepth vs EyeDetect:
44 Channels vs One

EyeDetect measures one physiological channel — the pupil — through a structured computer test. EchoDepth analyses 44 FACS-compliant facial Action Units per frame during any interaction. Here is the side-by-side.

EyeDetect, developed by Converus, is the most widely deployed alternative to polygraph in security screening contexts. It measures ocular-motor responses — pupil dilation, blink rate, reading patterns — during a structured computer-based test, arguing that deception produces measurable differences in these parameters. It is peer-reviewed, non-contact, and has been adopted in several law enforcement and government screening programmes internationally. EchoDepth is a different category of tool. Understanding the distinction matters for procurement decisions.

The Core Difference: One Channel vs Forty-Four

EyeDetect's measurement basis is pupillometry — specifically the observation that cognitive load and deception produce measurable changes in pupil diameter and eye movement patterns. This is a genuine physiological phenomenon supported by research. The limitation is that it is a single channel: pupil response is sensitive to lighting conditions, medications, fatigue, and many factors unrelated to deception.

EchoDepth analyses 44 FACS-compliant facial Action Units simultaneously — 44 distinct channels of facial muscle activation, each corresponding to a specific muscle or muscle group. AU combinations provide redundant, cross-validating evidence for emotional and cognitive states. A stress response that suppresses on one AU channel typically shows on several others. This multi-channel redundancy is significantly more robust than single-channel pupillometry for detection in adversarial conditions.

Capability EyeDetect EchoDepth
Measurement channels1 — pupillary response44 FACS Action Units simultaneously
Hardware requiredProprietary infrared eye-tracking camera + controlled lightingStandard RGB camera at 720p minimum — existing CCTV or webcam
Test protocolStructured computer test — subject must complete specific question sequenceAny interaction — interview, training, routine access — no structured protocol required
Cloud dependencyYes — results processed on Converus serversNone — fully on-premise, zero external transmission
SCIF / air-gap deployableNo — requires external connectivity for result processingYes — Docker on-premise, no outbound calls
UK data residency US company, cloud processingYes — default; all data in UK
Real-time continuous outputNo — session/test onlyYes — ~700ms latency, continuous AU stream
Works during natural interactionNo — requires dedicated test sessionYes — interview, review, access session
Countermeasure resistance Pupil suppression documented in literatureHigher — 44 channels, temporal sequencing, micro-expression detection
SIEM integrationNoYes — REST API, WebSocket, Splunk, Sentinel, QRadar
Insider threat / continuous vettingNo — point-in-time onlyYes — baseline + continuous anomaly scoring
UK-developed and hostedNo — US (Converus, Lehi Utah)Yes — Cavefish Ltd, Cardiff, Wales

When EyeDetect Is the Right Tool

EchoDepth's position is not that EyeDetect is without merit. It is a peer-reviewed, non-contact alternative to polygraph that measures a genuine physiological signal. In environments where a structured computer-based test protocol is operationally feasible, where external cloud connectivity is acceptable, and where the measurement of a single cognitive channel is sufficient, EyeDetect is a reasonable choice.

The specific use cases where EyeDetect works well include pre-employment screening in non-classified environments, customs and border screening programmes that can accommodate a structured protocol, and law enforcement applications in jurisdictions where it has been validated.

When EchoDepth Is the Right Tool

EchoDepth addresses a different set of operational requirements. For environments where SCIF or air-gap deployment is mandatory — EyeDetect cannot operate there. For continuous vetting and insider threat monitoring — EyeDetect's structured test protocol makes continuous monitoring operationally impractical. For integration with existing SIEM infrastructure — EyeDetect has no SIEM integration capability. For deployment on existing camera infrastructure without specialist hardware procurement — EyeDetect requires proprietary cameras.

For UK defence and intelligence procurement, the sovereignty question is also relevant: EyeDetect processes results on US-based servers. EchoDepth is developed, hosted, and processed entirely within the UK.

Both tools are also measuring different things. EyeDetect measures one channel of ocular-motor response during a structured test. EchoDepth measures 44 channels of facial muscle activation during any interaction, providing both point-in-time credibility assessment and continuous longitudinal monitoring. These are complementary capabilities, not equivalent alternatives.

Summary

The right question for procurement

If you need a structured, discrete credibility screening session in a non-classified, cloud-connected environment: EyeDetect is a viable option. If you need SCIF-compatible deployment, UK data residency, continuous vetting, SIEM integration, or monitoring during natural interactions rather than structured tests: those requirements point to EchoDepth.

Request a Procurement Briefing Credibility Assessment Overview