StarkByMed LogoStark
How It Works

Every Metric
That Matters

A single scan unlocks 8 clinical metrics, 30+ facial landmarks, golden ratio analysis, and a personalized treatment roadmap — all in under 3 seconds.

The Technology

Two AI Engines. One Report.

StarkByMed runs two independent AI pipelines simultaneously and fuses the results into a single clinical report.

Primary Engine

AWS Rekognition

  • 30+ facial landmark coordinates
  • Head pose — roll, yaw, pitch angles
  • Gaze direction per eye
  • Age range & gender estimation
  • 8 emotion confidence scores
  • Eye openness, smile, mouth state
  • Glasses, beard, face occlusion detection
  • Image brightness & sharpness quality

Supplemental Engine

Face++

  • Skin health score (0–100)
  • Acne severity index
  • Dark circle depth analysis
  • Skin staining & discoloration
  • Fine line & wrinkle detection
  • Non-blocking — never delays results

Runs in parallel with an 8s timeout — results are enriched when available, never blocked.

Clinical Output

8 Metrics Per Scan

Each metric is computed from raw landmark geometry and skin data — no self-reporting, no guesswork.

Symmetry Score

7 independent checks across eyes, brows, nostrils, mouth, jaw, and midline — corrected for head roll and yaw.

AWS Rekognition

Glow Score

Skin health, acne absence, fatigue level, and positive emotion confidence fused into a single radiance metric.

Rekognition + Face++

Skin Clarity

Weighted penalty model across acne, staining, dark circles, and fine lines blended with overall skin health.

Face++

Fatigue Index

Dark circles (40%), eye openness (30%), neutral emotion ratio (20%), skin dullness (10%).

Rekognition + Face++

Head Pose Score

Penalises yaw, roll, and pitch deviation from straight-on. Low scores trigger a retake prompt automatically.

AWS Rekognition

Photo Quality

Brightness (optimal 50–70) weighted 40% plus sharpness 60%. Flags poor captures before they corrupt results.

AWS Rekognition

Golden Ratio

4 proportion checks calibrated by detected gender — nose/mouth, pupil span, eye/nose, mid/lower face height.

AWS Rekognition

Pupil Distance

Inter-pupil distance as % of face width, displayed against the 42–50% aesthetic norm for immediate context.

AWS Rekognition

Symmetry Analysis

Not One Check.
Seven.

Most tools stop at eye alignment. StarkByMed runs seven independent geometric checks, each normalized by inter-ocular distance so scores are resolution-independent. Every measurement is corrected for head pose before scoring.

  • Eye vertical alignmentRoll-corrected
  • Facial midline offsetNose + chin average
  • Mouth tiltRoll-corrected
  • Eye size symmetryCorner-to-corner
  • Brow height symmetryRoll-corrected
  • Nostril width balanceRelative to nose tip
  • Jawline width balanceYaw-corrected

Symmetry Analysis

7 checks complete

82%

Pose-corrected symmetry

Eyes
96
Midline
98
Mouth
91
Brows
84
Jaw
78

Overall Score

70/100

Nose / Mouth Width
0.70 / 0.70
Pupil Span / Face Width
0.46 / 0.46
Eye Width / Nose Width
1.01 / 0.90
Mid / Lower Face Height
0.88 / 0.88

Proportion Analysis

Golden Ratio
Meets Clinical Data

Four facial proportions measured against research-backed aesthetic ideals — calibrated by detected gender, because the golden ratio isn't one-size-fits-all.

Gender-calibrated ideals

Male patients have proportionally wider noses and longer lower thirds — ideals adjust automatically.

Actual vs ideal ratio

Every check shows the measured ratio alongside the target so providers can discuss specific proportions.

Pose-independent scoring

Landmark positions are corrected for yaw before ratios are computed.

For Providers

The Complete Workflow

01

Share Your Scan Link

Every practice gets a unique URL and QR code from the settings page. Display it on a waiting room tablet or print it in intake forms.

02

Patient Scans In Seconds

No app download. Camera captures a selfie, two AI engines process it simultaneously. Full report in under 3 seconds.

03

Consult with Data

Patient sees their scored report. Your dashboard shows the full scan with every metric and AI-recommended treatment.

Provider Dashboard

Your Practice.
Your Analytics.

Every scan is saved to your account. Review full reports, track your top-recommended treatments, and manage your entire treatment catalog from one place.

  • Full scan history with all 8 metrics
  • Top recommended treatment tracking
  • Golden ratio + symmetry deep-dives per scan
  • Treatment catalog management
  • Practice settings + QR code generator
Go to Dashboard
Total Scans124
Avg Symmetry81%
Top TreatmentBotox
Patients Today8

For Patients

Data They Can Trust

The patient report shows personal scores, clinical notes, and a ranked treatment roadmap — all tied to specific measurements from their scan.

Personal Score Card

Four core metrics — symmetry, glow, skin clarity, and fatigue — with visual progress bars and qualitative labels.

Clinical Observations

Specific findings like 'Left brow sits higher' give patients language for what they've already noticed about their face.

Treatment Roadmap

Ranked recommendations from your practice's active catalog — each treatment tied directly to a measured metric.

Gaze & Pose Analysis

Head pose angles and gaze direction give context to every metric and show capture quality.

Skin Deep-Dive

Detailed skin health sub-scores: dark circles, staining, acne severity, and overall skin health index.

Proportion Report

The golden ratio breakdown shows each measured proportion versus the gender-calibrated ideal.

See It On
Your Own Face

Run a live scan right now — no account needed. All 8 metrics, golden ratio breakdown, and gaze analysis in under 3 seconds.

Start Free Scan