Ultra Fast Structure-aware Deep Lane Detection

By Zequn Qin, Huanyu Wang, Xi Li. European Conference on Computer Vision, 2020

547 citations recorded by the source corpus.

Modern methods mainly regard lane detection as a problem of pixel-wise segmentation, which is struggling to address the problem of challenging scenarios and speed. Inspired by human perception, the recognition of lanes under severe occlusion and extreme lighting conditions is mainly based on contextual and global information. Motivated by this observation, we propose a novel, simple, yet effective formulation aiming at extremely fast speed and challenging scenarios. Specifically, we treat the process of lane detection as a row-based selecting problem using global features. With the help of row-based selecting, our formulation could significantly reduce the computational cost. Using a large receptive field on global features, we could also handle the challenging scenarios. Moreover, based on the formulation, we also propose a structural loss to explicitly model the structure of lanes. Extensive experiments on two lane detection benchmark datasets show that our method could achieve the state-of-the-art performance in terms of both speed and accuracy. A light-weight version could even achieve 300+ frames per second with the same resolution, which is at least 4x faster than previous state-of-the-art methods. Our code will be made publicly available.

Subfields: Artificial Intelligence, Automotive Engineering, Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/24aa45720651dce230815710ce29a1714e2f5ab7

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Automotive Engineering

In Computer Vision and Pattern Recognition

Citations in the corpus

RESA: Recurrent Feature-Shift Aggregator for Lane Detection (2020, methodology): “For CULane dataset, several popular lane detection methods, including ResNet50 (Chen et al. 2017), ResNet101, SCNN, Res34-SAD (Hou et al. 2019), Res34-Ultra (Qin, Wang, and Li 2020) are used for comparison.”
CondLaneNet: a Top-to-down Lane Detection Framework Based on Conditional Convolution (2021, methodology): “On the one hand, the segmentation-based shape description is inefficient for lane lines due to the excessively high degree of freedom [30].”
CLRNet: Cross Layer Refinement Network for Lane Detection (2022, methodology): “Optionally, we can add an auxiliary segmentation loss following [19].”
Keep your Eyes on the Lane: Real-time Attention-guided Lane Detection (2020, methodology): “Compared to [20], our most lightweight model (with ResNet-18) surpasses their largest (with ResNet34) by almost 3% of F1 while being much faster (250 vs.”