CondLaneNet: a Top-to-down Lane Detection Framework Based on Conditional Convolution

Liu, Lizhe; Chen, Xiaohao; Zhu, Siyu; Tan, Ping

doi:10.1109/iccv48922.2021.00375

Cited by 174 publications

(95 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Instead of the popular top-down regime, [27] propose a bottom-up approach that focus on local details. [18] achieve state-of-the-art performance, but the complex conditional decoding of lane lines results in unstable runtime depending on the input image, which is not desirable for a real-time system.…”

Section: G Discussionmentioning

confidence: 99%

Rethinking Efficient Lane Detection via Curve Modeling

Feng¹,

Guo²,

Tan³

et al. 2022

Preprint

View full text Add to dashboard Cite

This paper presents a novel parametric curve-based method for lane detection in RGB images. Unlike stateof-the-art segmentation-based and point detection-based methods that typically require heuristics to either decode predictions or formulate a large sum of anchors, the curvebased methods can learn holistic lane representations naturally. To handle the optimization difficulties of existing polynomial curve methods, we propose to exploit the parametric Bézier curve due to its ease of computation, stability, and high freedom degrees of transformations. In addition, we propose the deformable convolution-based feature flip fusion, for exploiting the symmetry properties of lanes in driving scenes. The proposed method achieves a new state-ofthe-art performance on the popular LLAMAS benchmark. It also achieves favorable accuracy on the TuSimple and CULane datasets, while retaining both low latency (>150 FPS) and small model size (<10M). Our method can serve as a new baseline, to shed the light on the parametric curves modeling for lane detection. Codes of our model and PytorchAutoDrive: a unified framework for self-driving perception, are available at: https://github.com/ voldemortX/pytorch-auto-drive .

show abstract

Section: G Discussionmentioning

confidence: 99%

Rethinking Efficient Lane Detection via Curve Modeling

Feng¹,

Guo²,

Tan³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Moreover, the pixel-wise classification takes large computation resources. To overcome this, several work propose lightweight yet effective grid based (Qin et al, 2020;Liu et al, 2021a;Jayasinghe et al, 2021;Qu et al, 2021) or anchor based Li et al, 2019;Tabelini et al, 2021) methods. The grid-based approach detects lanes in a row-wise way, whose resolution is much lower than the segmentation map.…”

Section: D Lane Detectionmentioning

confidence: 99%

“…Modern Advanced Driver Assistance Systems (ADAS) for either L2 or L4 routes provide functionalities such as Automated Lane Centering (ALC) and Lane Departure Warning (LDW), where the essential need for perception is a lane detector to generate robust and generalizable lane lines (Comma.ai, 2017). With the prosperity of deep learning, lane detection algorithms in the 2D image space has achieved impressive results (Tabelini et al, 2021;Liu et al, 2021a;Qu et al, 2021), where the task is formulated as a 2D segmentation problem given front view (perspective) image as input (Lee et al, 2017;Pan et al, 2018;Neven et al, 2018;Abualsaud et al, 2021). However, such a framework to perform lane detection in the perspective view, 2D space is not applicable for industry-level products where complicated scenarios dominate.…”

Section: Introductionmentioning

confidence: 99%

PersFormer: 3D Lane Detection via Perspective Transformer and the OpenLane Benchmark

Chen¹,

Sima²,

Li³

et al. 2022

Preprint

View full text Add to dashboard Cite

Methods for 3D lane detection have been recently proposed to address the issue of inaccurate lane layouts in many autonomous driving scenarios (uphill/downhill, bump, etc.). Previous work struggled in complex cases due to their simple designs of the spatial transformation between front view and bird's eye view (BEV) and the lack of a realistic dataset. Towards these issues, we present PersFormer: an end-to-end monocular 3D lane detector with a novel Transformer-based spatial feature transformation module. Our model generates BEV features by attending to related front-view local regions with camera parameters as a reference. PersFormer adopts a unified 2D/3D anchor design and an auxiliary task to detect 2D/3D lanes simultaneously, enhancing the feature consistency and sharing the benefits of multi-task learning. Moreover, we release one of the first large-scale real-world 3D lane datasets, which is called OpenLane, with high-quality annotation and scenario diversity. OpenLane contains 200,000 frames, over 880,000 instance-level lanes, 14 lane categories, along with scene tags and the closed-in-path object annotations to encourage the development of lane detection and more industrial-related autonomous driving methods. We show that PersFormer significantly outperforms competitive baselines in the 3D lane detection task on our new OpenLane dataset as well as Apollo 3D Lane Synthetic dataset, and is also on par with state-of-the-art algorithms in the 2D task on OpenLane. The project page is available at https://github.com/OpenPerceptionX/OpenLane.

show abstract

“…Comparison on SDLane: It is challenging to detect highly curved lanes in the anchor-based detection framework, but the proposed algorithm provides excellent results on such curved lanes. To demonstrate this, on SDLane, we compare the proposed algorithm with the state-of-the-art techniques [19,30,34]. LaneATT [30] is an anchor-based method considering straight lines as anchors, while RESA [34] is based on the semantic segmentation framework.…”

Section: Comparative Assessmentmentioning

confidence: 99%

“…LaneATT [30] is an anchor-based method considering straight lines as anchors, while RESA [34] is based on the semantic segmentation framework. Recently, Cond-LaneNet [19] was proposed, which yields an F-measure of 79.48% on CULane. We train these methods on SDLane using the publicly available source codes.…”

Section: Comparative Assessmentmentioning

confidence: 99%

Eigenlanes: Data-Driven Lane Descriptors for Structurally Diverse Lanes

Jin¹,

Wonhui²,

Jeong³

et al. 2022

Preprint

View full text Add to dashboard Cite

A novel algorithm to detect road lanes in the eigenlane space is proposed in this paper. First, we introduce the notion of eigenlanes, which are data-driven descriptors for structurally diverse lanes, including curved, as well as straight, lanes. To obtain eigenlanes, we perform the best rank-M approximation of a lane matrix containing all lanes in a training set. Second, we generate a set of lane candidates by clustering the training lanes in the eigenlane space. Third, using the lane candidates, we determine an optimal set of lanes by developing an anchor-based detection network, called SIIC-Net. Experimental results demonstrate that the proposed algorithm provides excellent detection performance for structurally diverse lanes. Our codes are available at https://github.com/dongkwonjin/Eigenlanes.

show abstract

CondLaneNet: a Top-to-down Lane Detection Framework Based on Conditional Convolution

Cited by 174 publications

References 24 publications

Rethinking Efficient Lane Detection via Curve Modeling

Rethinking Efficient Lane Detection via Curve Modeling

PersFormer: 3D Lane Detection via Perspective Transformer and the OpenLane Benchmark

Eigenlanes: Data-Driven Lane Descriptors for Structurally Diverse Lanes

Contact Info

Product

Resources

About