Skip to content
This repository has been archived by the owner on Jun 2, 2021. It is now read-only.
/ DoodleNet Public archive

A web app which uses TensorflowJS and the webcam to play a game of Pictionary.

Notifications You must be signed in to change notification settings

kylecorry31/DoodleNet

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

1 Commit
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

DoodleNet

This application is designed to demonstrate the power of machine learning and computer vision using game mechanics. This application shows people the process of getting a machine learning model set up, as they have to train on different classes and will get the feel for how much training is needed for a simple drawing based application. They will then be able to play the games, which could also be used as a learning tool for people (especially kids) learning how to write letters or numbers. To make this app possible, innovations such as TensorFlow.js, TrackingJS, and FireBase were needed for the machine learning, computer vision, and persistence aspects respectively. This application addresses the problem as it allows users to see machine learning and computer vision in action, in a way that they get to actively participate with. As mentioned earlier, this project has the potential to help people learn to draw unfamiliar characters, given enough training data to the networks. To do this more effectively, it may need to be tweaked to not accept the answer in real time, as sometimes the network is sure you are drawing a specific symbol even if it is not finished.

Homescreen The homescreen.

Playing Playing the game.