king666888 / grit

Verifiable Goal Recognition for Autonomous Driving using Decision Trees

GRIT

This repo contains the implementation of the method described in the paper:

"GRIT: Fast, Interpretable, and Verifiable Goal Recognition with Learned Decision Trees for Autonomous Driving" by Brewitt, et al. [1] (IROS 2021)

In the paper described above, GRIT was compared to another method named IGP2, for which code is available here: https://github.com/uoe-agents/IGP2 [2]

Please cite:

If you use this code, please cite "GRIT: Fast, Interpretable, and Verifiable Goal Recognition with Learned Decision Trees for Autonomous Driving"

@inproceedings{brewitt2021grit,
  title={GRIT: Fast, Interpretable, and Verifiable Goal Recognition with Learned Decision Trees for Autonomous Driving},
  author={Cillian Brewitt and Balint Gyevnar and Samuel Garcin and Stefano V. Albrecht},
  booktitle={2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2021)},
  year={2021}
}

The files "evalutation/run_track_visualisation.py", "core/tracks_import.py", and "core/track_visualizer.py" are based on the inD Dataset Python Tools available at https://github.com/ika-rwth-aachen/drone-dataset-tools

#Setup Make sure you are using Python 3.6 or later.

Install Lanelet2 following the instructions here.

Clone this repository:

git clone https://github.com/uoe-agents/GRIT.git

Install with pip:

cd GRIT
pip install -e .

Extract the inD and rounD datasets into the GRIT/data directory.

Apply patches to the lanelet2 maps:

cd lanelet_map_patches
python patch_lanelet_maps.py

Preprocess the data and Extract features:

cd ../core
python data_processing.py

Train the decision trees:

cd ../decisiontree
python train_decision_tree.py

Calculate evaluation metrics on the test set:

cd ../evaluation/
python evaluate_models_from_features.py

Show animation of the dataset along with inferred goal probabilities:

python run_track_visualization.py --scenario heckstrasse --goal_recogniser trained_trees

References

[1] C. Brewitt, B. Gyvenar, S. Garcin, S. V. Albrecht, "GRIT: Fast, Interpretable, and Verifiable Goal Recognition with Learned Decision Trees for Autonomous Driving", in IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2021

[2] S. V. Albrecht, C. Brewitt, J. Wilhelm, B. Gyevnar, F. Eiras, M. Dobre, S. Ramamoorthy, "Interpretable Goal-based Prediction and Planning for Autonomous Driving", in Proc. of the IEEE International Conference on Robotics and Automation (ICRA), 2021