https://github.com/jithurjacob/tensorflowjs-gesture-controlled-2048

Play 2048 using Convolutional Neural network trained on your browser.
https://github.com/jithurjacob/tensorflowjs-gesture-controlled-2048

2048 gesture-recognition tensorflow tensorflow-experiments tensorflowjs

Last synced: 3 months ago
JSON representation

Play 2048 using Convolutional Neural network trained on your browser.

• Host: GitHub
• URL: https://github.com/jithurjacob/tensorflowjs-gesture-controlled-2048
• Owner: jithurjacob
• License: mit
• Created: 2019-01-04T12:53:12.000Z (almost 6 years ago)
• Default Branch: master
• Last Pushed: 2019-01-06T14:03:52.000Z (almost 6 years ago)
• Last Synced: 2024-09-27T21:03:45.072Z (3 months ago)
• Topics: 2048, gesture-recognition, tensorflow, tensorflow-experiments, tensorflowjs
• Language: JavaScript
• Homepage: https://jithurjacob.in/tensorflowjs-gesture-controlled-2048/UI/
• Size: 25.4 KB
• Stars: 9
• Watchers: 3
• Forks: 1
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# Gesture controlled 2048 - TensorFlow.js Transfer Learning

Play 2048 using Convolutional Neural network trained on your browser.

I've used TensorFlow.js to build a gesture control mechanism for playing 2048. It uses a pretrained MobileNet model and train another model using an internal mobilenet activation to predict the movements.

Try it [here](https://jithurjacob.in/tensorflowjs-gesture-controlled-2048/UI/)

# Video

[![YouTube](https://i.ytimg.com/vi/L0iSFf0VExQ/3.jpg)](https://youtu.be/L0iSFf0VExQ)