Skip to content

autonomous driving in gazebo with keras and ROS

Notifications You must be signed in to change notification settings

Exusai/EK_AutoNOMOS_Sim

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Self driving with ROS, AutoNOMOS Simulator, Gazebo, Keras y Tensorflow

This convolutional neural network is based on PilotNet but includes a set of convolutional layers dedicated to process information from a LIDAR sensor as an image:

Main.mp4

An implementation of PilotNet by Nvidia:

Comp.1.mp4

Neural Network Training Code: https://github.com/Exusai/EK_AutoNOMOS_Sim/blob/master/src/autonomos_gazebo_simulation/scripts/autodrive_evation_train.ipynb

Scripts used to train the network, includes manual drive to collect data and the notebooks used to train the network. Trainded models here.

About AutoNOMOS

Stand alone repo for the AutoNOMOS Gazebo simulation. Please check the wiki for information on how to use the simulation.

The files are divided in two:
* Gazebo plugin: Files to interact with the gazebo model.
* Autonomos: files of the model, world and example launch files.

Languages

  • Jupyter Notebook 90.9%
  • Python 6.5%
  • C++ 2.0%
  • Other 0.6%