#34 · Primary category: Robotics & Embodied AI Simulation

vision_pilot

adas advanced-driver-assistance-systems artificial-intelligence autonomous-driving autopilot autoware computer-vision deep-learning deep-neural-networks end-to-end-machine-learning foundation-models open-source robotics self-driving-car

Free and fully open-source L2 ADAS stack powered by End-to-End AI technology

Project last updated:08/25/26

GitHub Stars

874

Forks

185

Contributors

35

License

Apache-2.0

Why we included this project

Most open-source autonomous driving projects stop at perception demos or a collection of models, but Vision Pilot is built as something an OEM could realistically put into a production car. It targets entry-level L2 features such as adaptive cruise control, automatic emergency braking, lane keep assist, lane departure warning, and single-lane hands-free highway autopilot, the kind of functions Tier-1 suppliers integrate into passenger vehicles. The architecture pairs parallel perception models for safety with end-to-end models for driving behavior, so engineers can study how redundant checks and learned control are meant to work together. A single front-facing monocular camera and a mapless design keep hardware demands low and let it run where high-definition maps are unavailable. Model weights ship with the repo, and build routes from source, a Debian package, Docker images, ROS2 support, and CARLA simulation make the move from test-data evaluation to hands-on integration fairly direct.

Articles for this project

No articles for this project yet.

To suggest a topic or contribute an article, contact us.

Related projects in this category