RO-CBF

This repository contains the code needed to reproduce the examples found in L. Lindemann, A. Robey, L. Jiang, S. Das, S. Tu, N. Matni, Learning Robust Output Control Barrier Functions from Safe Expert Demonstrations

Setup the environment

conda create -n rocbf python=3.8.10
conda activate rocbf
pip install -r requirements.txt

Training Perception Map

python learning_cte.py

Training Perception-based ROCBF

bash launch_image_data.sh

Dependencies

• Carla 0.9.11 (we used overnight build version, but it should work for the full installation).