How victus-arc/face-detection works
This browser app loads a face-detection model, then lets the user start the camera and repeatedly detects faces in the video, drawing bounding boxes and confidence labels on a canvas. The principal runtime logic is all in script.js; the HTML and CSS provide its interface. The README also documents responsive layout and multiple-face detection. The graph follows the shown calls and README claims; no separate model or rendering implementation is present in the tree.
Main entry points: script.js Detection controller · index.html App page · styles.css Responsive styling
Components 8
Browser interface
- App page
index.html - Responsive styling
styles.css - Video and canvas
index.html
Detection runtime
- Detection controller
script.js - Detection overlay
script.js
Browser and model APIs
- Camera API
- face-api.js model
Other components
- User
Connections 9
- User → App page: uses
README.md - App page → Responsive styling: styled by
README.md - App page → Video and canvas: provides elements
- User → Detection controller: starts detection
script.js - Detection controller → Camera API: requests stream
script.js - Detection controller → Video and canvas: sets video source
script.js - Detection controller → face-api.js model: loads model and detects
script.js - Detection controller → Detection overlay: runs detection loop
script.js - Detection overlay → Video and canvas: draws boxes and labels
script.js