Magnifeye Labs
Myopia Detection
World-first AI technology that detects early signs of childhood myopia from photos.
The Story
See how this technology powers real-world applications in eye health screening.
Privacy & Security:
Your Photos are Secure
Secure Analysis
Photos are uploaded securely with encrypted local processing.
Instant Deletion
Photos permanently deleted after every analysis.
Not Stored
Images never leave the secure environment and are not stored.
Real-World Application
This technology powers Magnif-eye, a world-first tool that scans photos for early warning signs of myopia, screening for the eye condition years before most children are diagnosed.
468-Point Face Mesh
Our tech uses an advanced AI model to extract 478-point 3D meshes from a 2D photograph. These landmarks map the entire facial structure with remarkable precision.
Myopia Analysis
478
Landmarks
5176ms
Processing Time
88.1%
Confidence
85%
Symmetry Score
94.28px
Inter-Pupillary Distance
316
Mesh Triangles
3.6%
Face Coverage
960x1280
Resolution
Head Pose Estimation
Pitch
79.9°
Yaw
1.9°
Roll
1.3°
Face Bounding Box
Position
(83, 595)
Size
200x220
Key Landmark Coordinates
Multi-layered Technology
This face mesh technology is Layer 1 of Magnifeye's three-tiered AI model for detecting early signs of childhood myopia.
3D Facial Reconstruction
Generates a dense 3D geometric mesh from 2D photos, extracting ophthalmic and craniofacial parameters for analysis.
Medical Knowledge LLM
A multimodal LLM trained on pediatric myopia research conducts qualitative assessment of submitted photographs.
Behavioural Risk Model
Integrates environmental factors and family history as risk multipliers on top of image-based analysis.
These three signals are balanced within an aggregation layer to produce a risk classification, helping parents notice things they might have otherwise missed.
Clinical Validity
The technology and its knowledge-base is trained and based on clinical research in the identification of myopia through image-based photography.