[KongchangAI]
ProductFaceLandmarker

MediaPipe FaceLandmarker

An on-device face perception module in Google's MediaPipe family. It detects faces using BlazeFace and outputs 478 three-dimensional landmarks through a regression network, and can be compiled to WebAssembly to run in real time in the browser.

Timeline (last 90 days)

Oct 14

The FaceRuler app performs face inference on the browser side, using MediaPipe FaceLandmarker (478 keypoints) compiled to WASM, with a model of about 3.6 MB; the inference stage generates no third-party requests, and photos never leave the browser tab.

Unverified50%

All Facts (1)

Unverified

The FaceRuler app performs face inference on the browser side, using MediaPipe FaceLandmarker (478 keypoints) compiled to WASM, with a model of about 3.6 MB; the inference stage generates no third-party requests, and photos never leave the browser tab.

50%

Source Articles

Browser-Side Face Analysis in Practice: Building a Privacy-First Facial Assessment App with MediaPipeOct 14