Skip to content
No description, website, or topics provided.
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
assets
css
lib
README.md
index.html
webpack.config.js

README.md

mimic

make a face.

mimic is an in-browser game built using Affectiva facial recognition SDK.

Live

How to play:

Press start, and allow webcam access. Emoji will start falling from the top of the screen. Try to make a face that matches that emoji as closely as possible. Match as many as you can before time runs out. The more you get in a row, the higher the score multiplier.

Technical specs

Affectiva provides an incredibly powerful facial recognition tool that is able to detect various facial landmarks, which it uses to gauge expression and recognize emotion. It also provides an SDK which is loaded onto the local machine, crucial to providing the low latency needed for this game.

The game uses two canvases; one for the detector and webcam video, and another for the UI and falling emoji. When initialized, the detector creates a video and canvas element of its own, hiding the video element and feeding it to the visible canvas element. This is done so that facial landmarks can be painted on the canvas directly.

Detector & UI canvas

Diagram of callbacks

diagram

You can’t perform that action at this time.