Icon
Title
Try popping the falling drops with your body.
If your PC doesn’t have a camera, it won’t work.
When the web browser asks for access to the camera, please allow it. (Personal data never stored on the server.)
Please use a high-performance CPU for smooth image processing.
The web browsers currently working are as follows.
- Windows: Microsoft Edge, Google Chrome
- Android: Microsoft Edge, Google Chrome
- iOS: Safari
By placing icons and titles on the screen, you can create a special experience corner at your school or local science fair.
Javalab used 'PoseNet' as the human body detection library.
'PoseNet' is a vision model that can be used to predict the pose of a person in an image or video by predicting the positions of major body joints.
Posnet References
- TensorFlow: https://www.tensorflow.org/lite/examples/pose_estimation/overview?hl=ko
- GitHub: https://github.com/tensorflow/tfjs-models/tree/master/posenet
source code
We open source code to help you study.
https://editor.p5js.org/dongjoon9331@gmail.com/sketches/jevs3VZeC