AI Eye Bag Analyzer
No download required
Four Concerns, One Shadow
Almost nobody arriving with an under eye complaint knows which of four things they actually have. Puffiness is volume sitting forward. Dark circles are pigment or visible vasculature. A tear trough is a hollow casting a shadow. A droopy lower eyelid is position. All four present as tired eyes, all four are described the same way by the customer, and each one responds to something completely different. Scoring them separately is the whole point of the tool.
What the Eye Bag Analyzer Returns
Under eye puffiness on a 1 to 100 scale where higher means less visible, scored as its own value rather than folded into a general eye area figure.
Pigment and vasculature are different concerns with different answers, and conflating the two is the most common error in this category.
A hollow rather than a bulge, and the one most often mistaken for a bag because both create a shadow.
Eyelid position is anatomy rather than puffiness, and it is scored independently because nothing topical changes it.
The upper lid is scored separately from the lower one, because hooding changes how the whole eye area reads and is frequently blamed on the skin below it. A customer convinced their bags are worsening is sometimes watching the lid above descend instead.
The wider eye area is one of seven independently scored wrinkle regions, which separates fine lines from puffiness in a zone where the two are constantly confused. Crepey skin under the eye reads as a bag in a mirror and scores as a wrinkle here.
Dehydrated under-eye skin creases and reads as both puffier and darker than it is, so hydration belongs in the same response.
Region detections are returned as PNG images that can be overlaid on the original photo, so the customer sees which of the four concerns was actually flagged.
Measure Under Eye Puffiness in Three Steps
From selfie to routed recommendation in seconds, on any device the customer already owns.
Capture
A selfie at eye level in even indirect light, bare face, no flash, eyes open and relaxed. Overhead lighting casts a shadow that reads as both a bag and a dark circle, which is the single biggest source of a falsely poor score.
Analyse
The engine scores eye bags, dark circles, tear trough and eyelid position as four independent values, with moisture and texture alongside, returning structured data in seconds.
Route and Re-scan at Four Weeks
Puffiness driven by fluid and sleep can move within weeks, while structural change does not move at all, and the re-scan is what separates the two.
How to Route on the Eye Bag Score
Fluid or volume sitting forward, with no pigment component. Caffeine, cooling and sleep and salt habits are the realistic levers, and this customer can expect visible change in weeks.
Two concerns at once, and the pigment half will not respond to anything aimed at the swelling. Treat them as separate problems with separate products.
A hollow rather than a bulge. No cream fills a tear trough, and recommending one is how trust gets lost. This is a volume conversation with a practitioner.
Part of the appearance is dehydrated skin creasing and scattering light. Hydrate, repair and re-measure before recommending anything targeted.
Overhead Light Invents Eye Bags
Overhead Light Invents Eye Bags
The under eye is the area where lighting does the most damage to a measurement. A ceiling light casts a shadow directly into the orbital hollow, which reads as both puffiness and pigment at once, and a phone held below eye level adds a second shadow from the opposite direction. The same face measured under a bathroom spotlight at midnight and beside a window at noon can produce genuinely different scores, which is why the capture flow enforces even front lighting and eye level framing before the image is accepted rather than correcting for it afterwards.
Why Brands Choose Software Over Imaging Hardware
Imaging units typically run 5,000 to 10,000 US dollars each before servicing. Units are consumed per successful analysis instead, with volume based pricing.
One integration reaches your site, your app, a counter tablet and a clinic kiosk, rather than one location serving one customer at a time.
Under eye puffiness fluctuates with sleep and salt, so a single reading is weak evidence and a four week series is strong. That only works if re-scanning is effortless.
A hardware analyser that returns one eye area figure leaves the associate to interpret it. Four independent scores remove the guesswork.
Scores can be passed into your own systems with customer consent, turning a one off result into a segment you can market an eye care routine to.
Scientifically ValidatedChecked by Dermatologists
Checked by Dermatologists
The engine is trained on over 70,000 clinical images and reaches 95% test-retest reliability with over 80% correlation against professional physician assessment. A dermatologist-led review by Dr Steven R. Feldman, Professor of Dermatology at Wake Forest School of Medicine, examined its reliability, and the technology has been examined across four peer-reviewed journal publications. Validation covers the full Fitzpatrick range, which matters particularly here, because under-eye pigment presents differently across skin tones, and a system calibrated on one range will systematically misread another.
Who Uses an AI Eye Bag Analyzer
The under eye category is crowded with products aimed at four different problems, and a measured score decides which one to recommend.
A counter tablet that separates puffiness from pigment before the associate speaks, which prevents the single most common wrong recommendation in this aisle.
Tear trough and eyelid position are clinical concerns, and a structured intake score tells the practitioner which conversation to have before the consultation starts.
Puffiness is one of the few visible signs that responds to sleep and hydration within weeks, which makes it an unusually good proof point for a wellness product.
A single API that adds under eye scoring to an existing clinic, salon or ecommerce platform, returning structured data rather than a fixed interface.
Add Under Eye Scoring to Your Customer Experience
Talk to our team about integration, validation data by Fitzpatrick type, and deployment across web, app and in store.
Eye Bag Analyzer Frequently Asked Questions
What is an AI eye bag analyzer?
An AI eye bag analyzer estimates under eye puffiness from a photograph and returns a score from 1 to 100, where higher means less visible puffiness. It scores eye bags separately from dark circles, tear troughs and eyelid position, and it is a cosmetic assessment rather than a medical diagnosis.
What eye bag score do I need on the Perfect Corp analyzer?
Above 85 indicates minimal visible puffiness. The 70 to 84 band is mild, often fluid-related and responsive to habits; 40 to 69 is clearly visible and worth a targeted routine, and below 40 usually involves a structural component that topicals will not resolve. Scores are comparable within one system and over time for one person, not across vendors.
How accurate is the AI eye bag analysis?
The engine reaches 95% test to retest reliability and over 80% correlation with professional physician assessment, trained on 70,000+ clinical images and examined across four peer reviewed journal publications. Lighting affects the under eye more than any other region, which is why even front lighting is enforced at the point of capture.
Can beauty brands integrate under eye scoring into their own app or website?
Yes. Web and mobile SDKs, a server API and a hosted experience are all available, and the API returns structured data and detection masks rather than a fixed interface, so the presentation layer stays yours. Scores can be passed into your own CRM with customer consent.
Is the eye bag test free to use?
The hosted skin analysis experience can be tried free with no download. For commercial deployment, units are consumed per successful analysis with volume based pricing, rather than the 5,000 to 10,000 US dollars typical of imaging hardware. Contact sales for a quote against your expected volume.
