Dear authors:
I recently have reseached about lots of lane detection paper and modeling methods, but I found that most of them are based on roughly forward lanes rather than horizontal lanes. I think keypoint-based lane detection methods could be a solution. Should I segment the lane with v_samples instead of h_samples to get the keypoint labels?
Thanks!
Dear authors:
I recently have reseached about lots of lane detection paper and modeling methods, but I found that most of them are based on roughly forward lanes rather than horizontal lanes. I think keypoint-based lane detection methods could be a solution. Should I segment the lane with v_samples instead of h_samples to get the keypoint labels?
Thanks!