This project demonstrates an end-to-end learning technique which predicts steering angles of a car driving on urban roads. This is very helpful for the lane keeping operation.
The NVIDIA Neural Network architectre [1] is implemented in this code. The udacity dataset [2] is used for training and testing purposes. The whole code is written in a Google Colab Notebook and will require modifications if running on host computer.
Kindly refer to the attached report to checkout project details and results.I have also explained the code verbally in the attached video.
Project google drive can be viewed at - https://drive.google.com/drive/folders/1IUK3H5O7VNg2QGxeg6vBtlPGMde9SGkv?usp=sharing
[1] Source - https://arxiv.org/abs/1604.07316
[2] Source - https://github.com/udacity/self-driving-car/tree/master/datasets