Trust
Credibility you can inspect
A reading is only useful if you can see what produced it. This page states what the six attributes are, how each one is derived, what is deliberately not claimed, and how the data is handled.
The six attributes
The set is fixed. Adding attributes to make a report look fuller would make each one mean less.
Hydration
Surface moisture appearance.
Derived from how light scatters across the surface in the measured zones, compared against the same zones under the same framing conditions.
Texture
Smoothness and surface uniformity.
Derived from local variation in surface detail across the cheek and forehead zones.
Pores
Visible pore prominence.
Derived from the density and contrast of small surface openings in the zones where they are usually most visible.
Redness
Visible erythema distribution.
Derived from the distribution of red channel intensity relative to the surrounding tone, not from an absolute colour value.
Pigmentation
Tone evenness across zones.
Derived from tone difference between measured zones, so the reading is relative to your own skin rather than to anyone else's.
Firmness
Visible elasticity indicators.
Derived from surface appearance only: the visible indicators associated with elasticity in the lower face.
How a reading is scaled
Each attribute is expressed on a 0 to 100 scale, and each has a reference band, 60 to 80, where the reading sits comfortably. The band is used to order your report so the most actionable reading is first. There is no single overall score, because averaging six unrelated attributes into one number destroys the only useful thing about them.
What is deliberately not claimed
- No diagnosis, and no screening for any condition.
- No measurement of anything below the surface: no collagen, no barrier function.
- No prediction of how your skin will age.
- No before-and-after promises, and no timelines for visible change.
- No comparison of your skin against other people's readings.
How data is handled
The photograph
Used to produce the six readings, then not retained. No biometric template is kept, and nothing stored could reconstruct a face.
The readings
Six numbers and a date. They are what makes a later comparison possible, and they are the only thing an account needs to hold.
Consent
Given in plain language before capture, recorded with its date and scope, and withdrawable at any time from the account.
AI transparency
The readings are produced by a computer vision model, not by a clinician. A model reads an image under the conditions it was given: framing, lighting, camera and the moment of the day all move a reading, which is why the capture step checks them and why reassessment matters more than any single number.
Performance is not uniform across skin tones. Attributes that depend on colour distribution, redness and pigmentation in particular, are harder to read consistently on deeper skin tones, and a model that is not evaluated separately per tone group will hide that behind a single average. Any evaluation worth publishing has to report per-tone results, and until it does, a reading on deeper skin should be treated as less certain rather than presented with the same confidence.
A reading is a description of an image, not a fact about a person. Where the two disagree, the person is right.
PARTNER_LOGO_SLOT
Empty — no partner named in this demo
References
Empty — no studies cited in this demo
Both slots are left empty on purpose. Inventing a study, an institution or a partner logo to fill a layout would undo the point of this page.
Aicona Labs provides educational insights about visible skin attributes. It is not a medical device and does not diagnose, treat or prevent any condition.