Scholarlink

<p>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 <i>any</i> 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 <i>LaneCorrect</i>, 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.,&#xa0;<i>TuSimple</i>) detection without any human labels; (iv) We thoroughly evaluate our self-supervised method on four major lane detection benchmarks (including <i>TuSimple, CULane, CurveLanes</i> and <i>LLAMAS</i>) and demonstrate excellent performance compared with existing supervised counterpart, whilst showing more effective results on alleviating the domain gap, i.e.,&#xa0;training on <i>CULane</i> and test on <i>TuSimple</i>.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

LaneCorrect: Self-Supervised Lane Detection

  • 期刊论文

  • 发表日期：

    2025年3月24日

查看全文
  • Ming Nie,
  • Xinyue Cai,
  • Hang Xu,
  • Li Zhang

摘要

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.