Your Reality, Reimagined with Stickers !

Transform your reality with fun, interactive stickers. Express yourself like never before.

Launch Goggles AR Launch Moustache AR
AR Experience Example

Note: Optimised for desktop webcams only.

Implementation Details

P5.js

To display PoseNet's keypoints (nose, eyes, and skeleton points) as dots in P5.js, you get the `x` and `y` coordinates for each detected keypoint. Then, using P5.js's `circle()` or `ellipse()` functions, you draw a dot at those coordinates, customizing its appearance with `fill()` and `stroke()` if desired.

PoseNet.js

PoseNet.js is a browser-based machine learning model (built on top of TensorFlow.js) that performs real-time human pose estimation. It detects and tracks 17 key body joints in images or videos, enabling interactive applications that respond to human movement without sending data to a server.

ml5.js

ml5.js is an open-source JavaScript library that makes machine learning accessible for web creatives. Built on TensorFlow.js, it simplifies integrating ML models (like image classification, pose estimation, and text generation) directly into browsers.