The repo contains the codes of my experience with autonomous vehicles as part of the CS 598 MAAV: Methods for Building Autonomous Vehicles course, taught by Prof. David Forsyth at UIUC.
More details on assignments and project work can be found in their respective directories.
-
Homework 1: Object Detection - ImageNet
-
Challenge 1: Brake on Detecting Pedestrian
-
Challenge 2: Mapping and localization using Lidar
-
Challenge 3: Indoor and Outdoor SLAM
-
Project: Pedestrian Follower with PID control
-
Miscellaneous: Codes for controlling the AV using a keyboard, with and without PID control
All experiments are done on the vehicle platform by AutonomousStuff.
For development on my local machine, I installed Ubuntu 16.04 subsystem on my Windows 10 laptop. If you plan to do the same, then for RViz visualization/simulations, you may install XMing and/or VcXsrv.
Apart from this, we need to install ROS (Robot Operating System) that provides the necessary support to interact with the autonomous vehicle and for simulations as well. I worked with ROS Kinetic (Desktop-Full Installation). Install some basic ROS utils:
$ sudo apt-get install ros-kinetic-rviz
$ sudo apt-get install ros-kinetic-roscd
I am using Python 2.7 since found that its easily compatible with ROS Kinetic. Install OpenCV (3.3.1-dev):
$ sudo apt-get install python-opencv
For libraries used in the code, you should be able to install them using apt-get following the example. For python package pacmod_msgs used as,
import pacmod_msgsInstall using the following command:
$ sudo apt-get install ros-kinetic-pacmod-msgs
Note: The ROS packages you install can be found in /opt/ros/kinetic/. You might find the packages in share subdirectory within.
Thanks to our instructor, Prof. David Forsyth and my team members:
- Amber Gupta
- Bhavana Jain
- Devansh Shah
- Keval Morabia
- Manav Mehra
Happy to connect for discussions and details:
Jatin Arora
[email protected]
[email protected]