CLRNet: Cross Layer Refinement Network for Lane Detection

Benchmark Model Rank Results
lane-detection-on-llamasCLRNet (DLA-34)#1F1: 0.9612
lane-detection-on-llamasCLRNet (ResNet-18)#4F1: 0.9600
lane-detection-on-tusimpleCLRNet(ResNet-34)#1Accuracy: 96.9%F1 score: 97.82
lane-detection-on-tusimpleCLRNet(ResNet-18)#3Accuracy: 96.82%F1 score: 97.89
lane-detection-on-tusimpleCLRNet(ResNet-101)#28F1 score: 97.62
lane-detection-on-culaneCLRNet(DLA-34)#13F1 score: 80.47
lane-detection-on-culaneCLRNet(ResNet-101)#17F1 score: 80.13
lane-detection-on-culaneCLRNet(ResNet-34)#19F1 score: 79.73
lane-detection-on-culaneCLRNet(ResNet-18)#22F1 score: 79.58