Hand Written Digit Recognition

Overview: AI at the Edget to classify digit hangwriting.

Dataset: MNIST Hangwritten digits (28px x 28px grayscale 65,000 images)

Model: CNNs

Labels: 10 classes (the digits 0-9)



Environment: Git Static site development for AI at the Edge.



LIVE DEMO

Draw digit in gray area and click [Predict]


References:

🌐 Colab: TensorFlow.js