Comment by redox99
3 years ago
It does not use what's often called "HD maps". It uses vision + NNs to guess the lanes (even if they aren't visible), with some help of regular map data, but vision takes priority.
3 years ago
It does not use what's often called "HD maps". It uses vision + NNs to guess the lanes (even if they aren't visible), with some help of regular map data, but vision takes priority.
No comments yet
Contribute on Hacker News ↗