This is a neural network that does road segmentation. The code was written in Google colaboratory, but you can run it locally too.
This contains a jupyter notebook with useful operations for Google colaboratory. For example, mounting google drive, installing libraries, accesing file storage or displaying information about the hardware you use in colab.
Made a simple neural network that predicts numbers from mnist.It is implemented from scratch using only numpy, pandas and matplotlib. I followed the explanation from Samson Zhang (specially thanks) It uses RelU as activation function and softmax for clasifying the results.
This project uses canvas element and draws everything with JavaScript. In this project i used a neural network with genetic algorithm to learn how to drive. It starts with 100 cars, when the population is less than a procent it takes the best car that drived the most and mutates it.
This project let you draw a network of roads. This will be usefull when integrated with self driving car to create a city with autonomous drivings.