Choice before the camera
People read the data notice and choose a live scan, camera, or album. A single photo offers an alternative to moving their head.
A new way to read tradition
Hyean
From a scan to a reading,
in an understandable order.
A web and Apps-in-Toss project connecting face input with traditional Korean reading content.
Product · UX/UI · Engineering · Operations
Consent, input choice, preview, and detailed reading become one clear journey.
The starting problem
People need to know why a camera is requested and how their images are handled. Results also need a reading order. We designed input choices and result hierarchy together.
The experience
- Notice and consent
- Scan or select photo
- Preview a type
- Read and share
Inside the product
View original image ↗
Design decisions
From a preview to the detail
A type and summary come before the detailed reading. Sharing includes a card preview, and saved entries provide a route back.
Keep only the necessary input
The library does not retain photos or raw facial features. The server checks both the user and the exact reading scope before granting detailed access.
Implementation & scope
MediaPipe input is separate from shared reading contracts. Face, birth-chart, and palm journeys use distinct distribution profiles.
TypeScriptMediaPipeReactApps-in-Toss
A public web demo is available; Apps-in-Toss distribution is managed separately. Traditional interpretation is reference content, not medical assessment.
Open the service ↗