AI Driving Olympics

Tensorflow baseline "Solution template" for lane following challenge aido-LF

This is a baseline solution for one of the challenges in the the AI Driving Olympics.

The online description of this challenge is here.

For submitting, please follow the instructions available in the book.

Note: In particular you will need to set a token and specify your docker username before submitting.

Description

In this baseline template, driving behavior is learned using imitation learning from rosbag logs.

Follow the makefile in the individual folders of this repository.

1. Download and extract data using the Makefile in the extract_data folder.
2. Learn a model using the Makefile and scripts in the learning folder.
3. Submit the learned model by running make submit in the imitation_agent folder.

Each folder has its own README for further instructions.

This code has been tested on Mac OS and Ubuntu 16.04 both running Python 3.7 and Python 2.7.12.