Early access · staged rollout
Launching Nov 1, 2026
00Days
00Hrs
00Min
00Sec
AUREON LABS wordmark rendered as glowing point-cloud lettering on a black background

Face analysis that
shows its work.

Understand facial proportions, contour relationships, and profile geometry — each reading shown with its reference status and the limits of the photo behind it.

Phone number is optional — only used if we need to reach you about your access group.

or

Facial data is processed entirely in-memory and is never stored or sold.

You’re on the list. We’ll email you when early access opens.

Clear front and profile photos required · No payment details

Informational analysis. Not medical advice.

Point-cloud rendering of the skull and upper facial structure over fine calibration lines

Geometry layer — dense landmark field with symmetry calibration

WHAT MAKES IT DIFFERENT

A number can’t show its work.

A single rating hides how it was reached and what the photo could actually support. Instead of one verdict, each report shows the measured relationship, its reference status, the supporting landmarks, and the limits of the photo used to calculate it.

Your profile shows what a front photo cannot: how the nose and chin project, how the lips sit between them, and the overall facial relationship in side view.

Geometry first

Front and profile analyses use separate view-specific pipelines. Results that a view cannot support stay unavailable rather than being estimated from the other.

Confidence visible

Every reading carries a confidence tier and reference status. Exploratory and unsupported readings are labelled rather than folded quietly into an average.

Versioned by design

Each report records the engine version used to generate it. New modules are introduced with an explicit reference and confidence status, and may remain exploratory until their validation evidence is sufficient.

HOW IT WORKS

Three stages, visible evidence.

From first upload to ongoing tracking. Hover a step to watch it in action.

1

Upload process

Front and profile capture with pose and quality checks before analysis runs.

In-depth analysis 2
AI-produced guide 3
Progress tracking 4
Aureon Nexus 5

Not every photo supports every reading. Aureon Labs reports what the image supports — and tells you why when it cannot.

ANALYSIS DOMAINS

Separate domains, separate logic.

A face rendered as a cloud of connected nodes, illustrating landmark extraction
Landmarks in, structure out

Every photo is converted into a landmark map, then interpreted across separate analysis domains. Harmony and Angularity use their own measurement and reference logic. Shape Reference is reported descriptively, not as a rating. Some domains share the same front-view landmark and pose checks, so an unusable image can limit more than one front-view result.

Each image is screened for the pose and landmark conditions required by the analysis. Individual readings also carry a confidence tier based on their landmark construction and reference status. Some image conditions, including occlusion, extreme lighting, styling, and low resolution, can still limit what a photo supports.

01 · HARMONY

Proportion & balance

Scored structural relationships across front and profile geometry — canon proportions, alignment across facial thirds and fifths, and left-right symmetry.

02 · ANGULARITY

Contour definition

Reference-band contour analysis across jaw, cheek, nose, chin, and eye/brow regions, read from visible edges in the photo.

03 · SHAPE REFERENCE

Descriptive facial shape position

A descriptive shape-space comparison of global and regional facial landmark patterns against the selected reference distribution. Shown as a reference position, not a value judgment, identity claim, or rating contribution.

04 · SUPPORTING CONTEXT

Reported separately, never blended in

Image quality, visible hair framing, and future appearance modules are reported separately when the image supports them. They do not alter structural geometry results.

No composite verdict is produced from these domains, and no domain is weighted against another.

WHAT A REPORT SHOWS

See a real analysis.

A run of the Aureon Labs engine on a public figure’s photos — the same landmark overlay, harmony matrices, and metric ledger a member sees for their own reading.

Open full analysis in a new window ↗

PRO · PROGRESS TRACKING

See what changes, side by side.

Pro re-scans are captured under the same conditions each time, so this and later baselines line up cleanly — drag to compare.

After: the same subject with clearer skin and a groomed presentation
Before: the same subject's initial baseline photo

Illustrative comparison — grooming, skincare and presentation changes shown; not a guaranteed or typical outcome.

CONFIDENCE & LIMITS

Every reading is labelled before you see it.

Each measurement is assigned one of four states based on its landmark construction, reference status, and what the photo supports.

  • Compared against measured ranges from a reference population.

  • Computed, but its target range is still being validated — treat as directional.

  • Pose, lighting, or landmark confidence reduced the certainty of this reading.

  • This image or the current landmark system can’t support this measurement.

What Aureon Labs is not.

  • Informational image-based analysis only — not medical advice, diagnosis, treatment, or a medical device.
  • Not a universal attractiveness score, and not a guarantee of any appearance outcome.
  • Not a direct observation of bone or soft tissue — readings describe visible landmark and contour patterns in a 2D photo.
  • Pose, lens, lighting, expression, and landmark quality affect what a photo supports.
  • Some readings are exploratory or unavailable under the current landmark system.
  • Results should not be treated as clinical measurements.

METHODOLOGY

Current validation status.

Validation is ongoing and reported per domain rather than as a single headline figure. Numbers are published only alongside the protocol behind them.

Front landmark geometry

Validated internally against reference-image fixtures. Metric reliability varies by landmark and construction.

Profile geometry

Uses a separate detector and side-pose gate. More profile measurements remain proxy-based or unavailable.

Reference bands

Some are externally sourced; others are internal reference distributions or exploratory ranges. Each report identifies the applicable status.

Illustrative confidence topology — a brand visual, not measured validation data.

Full methodology and validation notes will be published alongside the first release. Independent third-party validation has not been carried out.

EARLY ACCESS

Join the early-access list.

Early access opens in small groups while we validate new modules and incorporate participant feedback. Here is exactly what joining does:

  • An invitation by email when your access group opens
  • New modules and explanations as releases ship
  • No payment details required, and no subscription is started
  • A chance to give feedback before wider release

The waitlist does not collect photos. Photo handling, retention, and deletion controls will be published before analysis access opens.

FAQ

Worth knowing before you join.

What is Aureon Labs?

An evolving facial-geometry and appearance-analysis platform. It uses clear front and profile photos to organize visible structural relationships and explain each reading in plain language, with its confidence tier and reference status.

Why do you need both a front and a profile photo?

They show different things. The front view carries proportion, balance, and symmetry. The profile shows how the nose and chin project, how the lips sit between them, and the overall facial relationship in side view. The two views run through separate pipelines.

Is this medical advice?

No. Aureon Labs provides informational image-based analysis only. It is not a medical device, diagnosis, or substitute for advice from a qualified professional.

What photos are required?

A clear front-facing photo and a clear side-profile photo, in good lighting with minimal pose distortion. Aureon Labs indicates when photo quality limits a reading — not every photo supports every reading.

What happens to my photos?

The waitlist collects no photos at all. Photo handling, retention, deletion controls, and whether derived geometry is kept will be published in full before analysis access opens.

What happens after I join?

Your email joins the early-access list. Access opens in small groups; when your group opens you receive an invitation by email. Pricing has not been announced, and joining starts no subscription.

Are all readings equally validated?

No. Aureon Labs distinguishes reference-backed readings from exploratory ones, labels each accordingly, and marks a reading unavailable when the photo or current landmark system cannot support it.

QUESTIONS

Something not covered above?

Reach us at support@aureonmetrics.com.