Personalized ADAS in CARLA

This project aims to personalize Advanced Driving-assistance Systems (ADAS) for autonomous vehicles using the self-driving car simulator CARLA. The two main focuses are on personalizing lane changing and personalizing Adaptive Cruise Control (ACC) that includes vehicle following and lane following.

Hardware-In-Loop based simulation with torque-controlled steering motor and CARLA.

Learning Phase

The Learning Phase uses a Gaussian Mixture Model (GMM) for clustering user preferences into known presets that the ADAS can act on.

The vehicle starts in autopilot mode. Press p to toggle between manual control and autopilot.

In manual control mode:

Press l to start or end a learning recording session (collects driver behavior data).
Press t to train the model on the collected data and save it locally.

After training, press Backspace to reset the scene, then enable autopilot to observe the personalized behavior.

It is advisable to personalize lane following in scene 0, clone vehicle following behavior in scene 1, and teach lane changing in scene 2.