Scientific Methodology & Privacy Architecture
Detect Face Shape combines clinical craniofacial anthropometry with client-side computer vision to provide objective facial geometry classification.
Zero-Server Biometric Privacy Guarantee
Pure client-side execution via WebAssembly & WebGL
Most face analysis web tools require users to upload high-resolution photos to a remote cloud server where neural networks process and store them. This poses substantial privacy risks and biometric security concerns.
detectfaceshape.app operates with a zero-server architecture:
- The Google MediaPipe Face Landmarker model binaries and WASM runtime execute strictly within your local browser sandbox.
- All uploaded photos and camera frames are decoded onto an offscreen canvas in local browser volatile memory and deleted immediately after classification.
- No image data, biometric coordinates, or vector representations are ever sent across a network cable. You can even disconnect your internet once the page has loaded, and the detector will continue to function flawlessly.
Craniofacial Anthropometry & Key Landmarks
Human facial classification in detectfaceshape.app is rooted in standardised medical anthropometry pioneered by Leslie G. Farkas. Rather than guessing, our algorithm extracts 4 precise physical dimensions from 468 discrete landmark coordinates:
Trichion (10) to Menton (152)
The vertical distance between the centre top of the hairline and the bottom apex of the chin, defining overall facial elongation.
Zygion Left (234) to Zygion Right (454)
The maximum horizontal distance between the lateral curves of the zygomatic arches, establishing the primary horizontal axis.
Frontotemporale Left (127) to Right (356)
The horizontal breadth between the temples across the forehead, critical for identifying Heart and Triangle geometries.
Gonion Left (58) to Gonion Right (288)
The distance across the bilateral gonial angles of the mandible, diagnostic of Square and Triangle angularities.
Academic References & Citations
- Farkas, L. G. (1994). Anthropometry of the Head and Face (2nd ed.). New York: Raven Press. Standardised measurements of facial proportions and normative craniofacial values across diverse populations.
- Liu, Z., Lin, Y., Cao, Y., Hu, H., Wei, Y., Zhang, Z., Lin, S., & Guo, B. (2021). Swin Transformer: Hierarchical Vision Transformer using Shifted Windows. Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), 10012–10022.
- Lugaresi, C., et al. (2019). MediaPipe: A Framework for Building Perception Pipelines. arXiv preprint arXiv:1906.08172. Google Research.