Overview

• This repository provides the source code to reproduce the results of the Google Smartphone Decimeter Challenge 2023 on Kaggle
• Decimeter accurate position estimation was achieved using smartphone GNSS and IMU
  • Public score: 0.789 m (1st)
  • Private score: 0.928 m (2nd)
• Factor Graph Optimization of GNSS and IMU for Smartphones
• For details of the method, please refer to the following papers, presentations, and source code

Paper/Presentation

Taro Suzuki, "An Open-Source Factor Graph Optimization Package for GNSS and IMU Integration in Smartphones," in Proceedings of the 37th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2024), 2024.

• Paper
• Presentation

Test environments

• Ubuntu 20.04
• MATLAB R2024a
  • (Optional) Parallel Computing Toolbox

Requirements

• GTSAM: Factor graph optimization library. Due to a problem with the MATLAB wrapper, it is necessary to build GTSAM version 4.0.3 by following steps. Please refer to the GTSAM github for details on the build procedure.

sudo apt-get install build-essential cmake libboost-all-dev libtbb-dev
git clone https://github.com/borglab/gtsam.git
cd gtsam
git checkout 4.0.3
mkdir build
cd build
cmake .. -DCMAKE_BUILD_TYPE=Release -DGTSAM_BUILD_UNSTABLE:OPTION=OFF -DGTSAM_BUILD_EXAMPLES_ALWAYS:OPTION=OFF -DGTSAM_INSTALL_MATLAB_TOOLBOX:OPTION=ON
make
sudo make install

• MatRTKLIB: MATLAB wrapper for RTKLIB.

git clone https://github.com/taroz/MatRTKLIB.git

Add the MatRTKLIB installation directory to the MATLAB search path.

• MatlabProgressBar: tqdm like MATLAB progress bar.

git clone https://github.com/JAAdrian/MatlabProgressBar.git

Add the MatlabProgressBar installation directory to the MATLAB search path.

How to run

1. Install all of the above dependent packages
2. Clone this repository

git clone https://github.com/taroz/gsdc2023.git

3. Download the pre-processed GSDC2023 data set (2.7GB)

cd gsdc2023
wget http://www.taroz.net/data/dataset_2023.zip
unzip dataset_2023.zip

4. Run run_fgo.m in MATLAB
5. The submission file for Kaggle is generated in results directory. If you have Kaggle account, let's submit the estimation result to the Google Smartphone Decimeter Challenge 2023 and evaluate the accuracy. You should get the following results.