BodyFaceSoul
A computer vision layer that translates a photograph into structured signals and a visually explainable result.
Context
The mobile product needs a reliable path from source image to interpretation: detect landmarks, calculate geometry, render annotations and pass structured data to AI without losing context between services.
How it works
Separated processing into focused APIs: MediaPipe handles landmarks, a dedicated layer computes geometry and renders annotations, and AI produces the textual interpretation. Cloud Run, Firebase Functions, storage and the mobile interface operate as one asynchronous workflow.
What it enables
The result is a modular visual-analysis system: each stage can be validated independently while the app receives structured signals, an annotated image and a clear textual interpretation.