- Real-vehicle data and simulated data for testing
- Consume real-vehicle sensor data for testing
- Consume synthetic image data from CARLA simulation tool to efficiently mimic different real-world conditions
- Use rosbags when appriate with vehicle sensor data to emulate vehicle integration.
- Annotated AI perception output
- Real-vehicle control based on AI perception output (work in progress)
1. Clone Git:
1.1: git clone https://github.com/usnistgov/NIST_AV_AI.git cv_app
2. Install dependecies:
2.1: pip install ultralytics
3. Change into the ros working directory
4. Running Publisher:
4.1: colcon build --packages-select cv_app
4.2: source install/local_setup.bash
4.3: ros2 run cv_app sensor_publisher
5. Running Subscriber
5.1: source install/local_setup.bash
5.2: ros2 run cv_app sensor_subscriber