# MOM - 03.02.2023
- Plan:
- design a simple scenario (car in 4 different adverse conditions: day/night + rain/no rain)
- each car has its own sensor set (camera only, camera + lidar)
- limit the different classes set (3, 4)
- Get ground truth for (position + bounding boxes) from CARLA
- 1 data generator
- Compare (position + bounding boxes) from detection model against ground truth in all adverse conditions scenarios (performance metric: mAP)
- Tasks:
- Ahmad:
- provide the detection model (camera)
- search for the proper detection model (multi-modal)
- Mostafa:
- Build simple scenario in CARLA with sensors & stuff (Python) (starting point: visualize_multiple_sensors.py in PythonAPI/Examples)

- Workaround:
- Impose different noise matrices for the adverse conditions in case of no proper detection model(s)

- important docs:
- paper: https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=9551968