Skip to content

Experiment to re-create character recognition algorithm on par with the Apple Watch message app

Notifications You must be signed in to change notification settings

mwhitt/LetterDraw

Folders and files

NameName
Last commit message
Last commit date

Latest commit

author
Matthew Whittaker
Oct 13, 2019
46ae34e · Oct 13, 2019

History

15 Commits
Oct 13, 2019
Oct 13, 2019
Oct 13, 2019
Oct 13, 2019
Oct 13, 2019

Repository files navigation

Hand drawn letter recognizer

Experiment to recreate a hand drawn, single character ML model on par with the Apple Watch app.

Setup Xcode

Download the latest iOS opencv framework into the LetterDraw folder at the root of the project: https://opencv.org/releases/

Setup Training Model

Setup a python virtual environment:
https://medium.com/@MattGosden/set-up-your-mac-for-python-and-jupyter-using-virtual-environments-730bf2888e05

Install requirements:
pip install keras==2.2.4 tensorflow==1.14.0 coremltools

Download emnist-letters-test.csv & emnist-letters-train.csv into the data directory:
https://www.kaggle.com/crawford/emnist

About

Experiment to re-create character recognition algorithm on par with the Apple Watch message app

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published