🤖 AI Summary
To address the challenge of unsupervised lane detection in autonomous driving under label-free scenarios, this paper proposes the first LiDAR-intensity-driven framework for unsupervised 3D lane segmentation and self-correcting 2D label generation. Methodologically, it leverages LiDAR point cloud intensity to reconstruct 3D lane geometry, followed by geometrically consistent projection to yield initial 2D pseudo-labels; a novel self-supervised paradigm, LaneCorrect, is introduced, integrating instance-aware adversarial augmentation, geometric consistency constraints, and knowledge distillation for end-to-end pseudo-label refinement. Key contributions include: (i) pioneering a LiDAR-intensity-guided unsupervised 3D→2D lane modeling pipeline that substantially mitigates domain shift; and (ii) achieving performance on par with fully supervised methods across four major benchmarks—TuSimple, CULane, CurveLanes, and LLAMAS—with notable cross-domain (CULane→TuSimple) detection accuracy gains.
📝 Abstract
Lane detection has evolved highly functional autonomous driving system to understand driving scenes even under complex environments. In this paper, we work towards developing a generalized computer vision system able to detect lanes without using any annotation. We make the following contributions: (i) We illustrate how to perform unsupervised 3D lane segmentation by leveraging the distinctive intensity of lanes on the LiDAR point cloud frames, and then obtain the noisy lane labels in the 2D plane by projecting the 3D points; (ii) We propose a novel self-supervised training scheme, dubbed LaneCorrect, that automatically corrects the lane label by learning geometric consistency and instance awareness from the adversarial augmentations; (iii) With the self-supervised pre-trained model, we distill to train a student network for arbitrary target lane (e.g., TuSimple) detection without any human labels; (iv) We thoroughly evaluate our self-supervised method on four major lane detection benchmarks (including TuSimple, CULane, CurveLanes and LLAMAS) and demonstrate excellent performance compared with existing supervised counterpart, whilst showing more effective results on alleviating the domain gap, i.e., training on CULane and test on TuSimple.