论文标题
超级：新型车道检测系统
SUPER: A Novel Lane Detection System
论文作者
论文摘要
在过去的几年中，对基于AI的车道检测算法进行了积极研究。与传统的基于功能的方法相比，许多人表现出了出色的性能。但是，准确性通常仍处于低80％或高90％的位置，或者在使用具有挑战性的图像时降低。在本文中，我们提出了一个实时车道检测系统，称为场景理解物理增强的实时（超级）算法。所提出的方法由两个主要模块组成：1）作为场景提取器的分层语义分割网络和2）物理增强了泳道推理的多车道参数优化模块。我们使用来自CityScapes，Vistas和Apollo的异质数据训练拟议系统，并评估四个完全独立的数据集（从未见过的）上的性能，包括Tusimple，Caltech，Caltech，Urban Kitti-Road和X-3000。所提出的方法的性能比已经在同一数据集上训练的车道检测模型相同或更好，即使在从未经过培训的数据集上也表现良好。还进行了实际的车辆测试。初步测试结果表明，与Mobileye相比，实时车道检测性能有希望。
AI-based lane detection algorithms were actively studied over the last few years. Many have demonstrated superior performance compared with traditional feature-based methods. The accuracy, however, is still generally in the low 80% or high 90%, or even lower when challenging images are used. In this paper, we propose a real-time lane detection system, called Scene Understanding Physics-Enhanced Real-time (SUPER) algorithm. The proposed method consists of two main modules: 1) a hierarchical semantic segmentation network as the scene feature extractor and 2) a physics enhanced multi-lane parameter optimization module for lane inference. We train the proposed system using heterogeneous data from Cityscapes, Vistas and Apollo, and evaluate the performance on four completely separate datasets (that were never seen before), including Tusimple, Caltech, URBAN KITTI-ROAD, and X-3000. The proposed approach performs the same or better than lane detection models already trained on the same dataset and performs well even on datasets it was never trained on. Real-world vehicle tests were also conducted. Preliminary test results show promising real-time lane-detection performance compared with the Mobileye.
