/posemoji

App using Tensorflow.js and HTML canvas

Primary LanguageTypeScript

Posemoji

App using Tensorflow.js to analyse poses using the users webcam

Commands

Run project

npm run start

Build project

npm run build

Requirements

  • Use Media Devices API to access a camera.
  • Use TensorFlow.js with a pre-trained model to detect human poses.
  • Choose both emojis and body parts freely.
  • The app should scale content to fit a window.
  • Host the app on a free hosting under a public domain, e.g. GitHub Pages or Firebase Hosting.
  • There must be a single button to switch between all available cameras.
  • The app should work on the newest stable version of a Chrome browser, both mobile and desktop.

Optional refactoring

  • Use async await instead of promises