search

voldemortX / pytorch-auto-drive

PytorchAutoDrive: Segmentation models (ERFNet, ENet, DeepLab, FCN...) and Lane detection models (SCNN, RESA, LSTR, LaneATT, BézierLaneNet...) based on PyTorch with fast training, visualization, benchmarking & deployment help
BSD 3-Clause "New" or "Revised" License
837 stars 137 forks source link

Why instance？ #120

Closed champagne-yellow closed 1 year ago

champagne-yellow commented 1 year ago

Nice work! Why does lane detection need instance segmentation? Is semantic segmentation not applicable?

voldemortX commented 1 year ago

@champagne-yellow Semantic segmentation can also work (In this framework: Baseline, SCNN, RESA), but in a restrictive way, i.e., restricting the max number of lanes by giving them semantic classes. 1 for ego left, 2 for ego right, etc.

voldemortX commented 1 year ago

This issue seems to be addressed. I'll close it for now. Feel free to continue commenting for reopen/open a new one.