Skip to main content

Overview

GlamAR’s AI Facial Skin Analysis uses deep learning and computer vision to evaluate skin health in real-time. With just one camera capture, it can detect 14+ skin conditions β€” from acne and wrinkles to pigmentation and pores β€” and return visualized, annotated results.

No wearables or external hardware required.

What Is It For?​

  • Provide personalized skincare insights to customers
  • Power AI-driven product recommendations
  • Enable skin assessments in apps, kiosks, or smart mirrors
  • Support dermatology-lite solutions for beauty, wellness, and medtech

What It Detects​

ParameterCategoryTypeRange / ValuesAnnotated ImageNotes
Skin ScoreOverallInteger0–100βœ…Overall score based on all sub-metrics
Skin TypeClassificationClassoily, dry, normal, combination❌Based on U-zone and T-zone variation
Skin ToneClassificationClassfair, light, medium, olive, tan, deep❌Melanin-based classification
Skin AgeClassificationInteger18–80❌Perceived age in years
AcneKey ConditionInteger0–100βœ…Whiteheads and inflamed pimples detection
WrinklesKey ConditionInteger0–100βœ…Fine lines and deep fold estimation
PoresKey ConditionInteger0–100βœ…Visibility and density of pores
Eye BagsEye AreaInteger0–100βœ…Lower eyelid puffiness
Under EyeEye AreaInteger0–100βœ…Orbital fat changes
Dark CirclesEye AreaInteger0–100βœ…Hyperpigmentation under the eyes
ScarsSpecific IssueInteger0–100βœ…Depth and presence of scars
WhiteheadsSpecific IssueInteger0–100βœ…Closed comedones evaluation
PigmentationSpecific IssueInteger0–100βœ…Local melanin concentration
HydrationSpecific IssueInteger0–100βœ…In development
RednessSpecific IssueInteger0–100βœ…In development
Projected Skin (1)ProjectionIntegerN/Aβœ…Filtered forecast: 1-month outcome
Projected Skin (2)ProjectionIntegerN/Aβœ…Filtered forecast: 3-month outcome

Integration Options​

MethodUse Case
SDKsFor real-time camera capture and instant analysis (Web, Android, iOS)
APIsFor backend-based analysis using image input
PluginsShopify, Webflow, Mirror UI modules
ConsoleUpload datasets, test reports, view logs

πŸ“¦ SDK is ideal for real-time camera mode.

πŸ“‘ API is best for batch or backend-triggered scans.

How It Works (Under the Hood)​

  1. Image Capture – via live camera or uploaded image
  2. ML Inference – Face is analyzed using trained CNN models
  3. Result Generation – 14+ skin metrics + annotated photo
  4. Visual Feedback – Returned to client as JSON with image

Models are optimized to run in under 3 seconds on-device or .5s via cloud.