Skip to content

Kladar/self-driving-car-nanodegree-part-2

Repository files navigation

This is a machine learning pipeline based on LeNet that can classify street signs in the data set at 99% accuracy.

It was trained on an AWS GPU in Oregon, and I watched it train with my laptop tethered to my LTE WIFI hotspot in rural central Wisconsin. The future is here and it's amazing.

Machine Learning is still magic to me.