Skip to content

Latest commit

 

History

History
18 lines (11 loc) · 708 Bytes

File metadata and controls

18 lines (11 loc) · 708 Bytes

Handwriting Recognition Project for learning

Special thanks to Mr. Johan Dettmar. Please come and find this awesome for details, he give clear instruction about how to do it.

I am using Tensorflow.js for creating the model, after training in colab based on EMNIST dataset. Find the jupiter notebook for your details.

Let's have fun at : https://tommy-ngx.github.io/hand_words/index.html

Model is buiding in Keras using some kernels, adam for optimizers, loss based on categorical_crossentropy. All layers are below:

accessibility text

Cheers,

Tommy Ngx