search

cardwing / Codes-for-Lane-Detection

Learning Lightweight Lane Detection CNNs by Self Attention Distillation (ICCV 2019)
MIT License
1.04k stars 333 forks source link

Which paper is ERFNet-CULane-PyTorch based on? #321

Closed nisha1729 closed 3 years ago

nisha1729 commented 3 years ago

Hi, could you tell which paper you used for the codes in ERFNet-CULane-Pytorch? I believe the underlying model is this: ERFNet: Efficient Residual Factorized ConvNet for Real-time Semantic Segmentation. Can you explain how you modified it for lane detection? Thanks!

cardwing commented 3 years ago

Basically, we take the ERFNet as the segmentation backbone(Encoder + Decoder) and add the lane existence prediction branch following the original SCNN paper. More details can be found in the model definition of ERFNet-CULane-PyTorch. Note that some preprocessing operations and training strategies also matter.

nisha1729 commented 3 years ago

Thank you!