Fiend_or_Foe_SLAM

Authors: Muhammad Haris Ikram, Saran Khaliq, Muhammad Latif Latif Anjum, Wajahat Hussain

Friend_or_Foe SLAM is modified version of ORB-SLAM with inclusion of adversarial database to detect and prevent false loop closures. Whenever a loop closure is detected by ORB-SLAM, the loop closing candidate is matched with adversarial database. If the match is found with adversarial database, the pipeline is terminated and loop closure is prevented. If the match is not found with adversarial database, the Freind_or_Foe SLAM works as ordinary ORB-SLAM.

#1. Prerequisites (ORB-SLAM Installation)

In order to run Friend-or-Foe SLAM, you will have to successfully install and run the ORB-SLAM. The github link for installation and running ORB-SLAM is given below:

	https://github.com/raulmur/ORB_SLAM

#2. Modifying ORB-SLAM

1. Make sure that ORB-SLAM is running correctly.

2. Clone the repository:

    git clone https://github.com/hkhanr/Friend_or_Foe_SLAM.git
   

3. Navigate to Friend_or_Foe_SLAM directory and do the following:

Copy the "Adv_database" and "Deciever_code" folders and paste them in ORB_SLAM folder. Also replace the files in "include" directory of ORB-SLAM with that of Friend_or_Foe_SLAM directory. In ORB_SLAM root directory, open "CMakeLists.txt" and do the following:

Replace the line

	src/LoopClosing.cc

by following line

	#src/LoopClosing.cc

Paste the following line just below it

	Deciever_code/LoopClosing.cc

All the necessary modification are done. Now we need to build the ORB-SLAM again.

#3. Installation

1. Build ORB_SLAM. Go into the ORB_SLAM root and execute the following:

    cd build
    cmake ..
    make
   

#4. Run SLAM

Launch the SLAM using following command:

	roslaunch ExampleGroovyOrNewer.launch

#5. Example Bag Files

Instrucions for running an example bag file on SLAM are given in the following link

	https://github.com/raulmur/ORB_SLAM

Our example bag files can be download from the following link

	https://drive.google.com/drive/folders/17o6nZJU3RpP6oHx6aoihJ07kfceP3NGU?usp=sharing

Also the camera setting file for our example bag can be downloaded from above link and the default setting file should be replaced by it.

If you use this repo, consider citing the corresponding paper:

@article{ikram2022perceptual,

title={Perceptual Aliasing++: Adversarial Attack for Visual SLAM Front-End and Back-End},

author={Ikram, Muhammad Haris and Khaliq, Saran and Anjum, Muhammad Latif and Hussain, Wajahat},

journal={IEEE Robotics and Automation Letters},

volume={7},

number={2},

pages={4670--4677},

year={2022},

publisher={IEEE} }