Home

Awesome

PersFormer: a New Baseline for 3D Laneline Detection

pipeline

PersFormer: 3D Lane Detection via Perspective Transformer and the OpenLane Benchmark
Li Chen<sup>∗†</sup>, Chonghao Sima<sup></sup>, Yang Li<sup></sup>, Zehan Zheng, Jiajie Xu, Xiangwei Geng, Hongyang Li<sup></sup>, Conghui He, Jianping Shi, Yu Qiao, Junchi Yan. <sup></sup> equal contributions. <sup></sup> corresponding authors

Introduction

PWC

This repository is the PyTorch implementation for PersFormer.

PersFormer is 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. It 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.

Changelog

Get Started

Installation

Dataset

Training and evaluation

Benchmark

MethodVersionAllUp &<br>DownCurveExtreme<br>WeatherNightIntersectionMerge&<br>SplitBest modelx-cx-fz-cz-fCategory Accuracy
GenLaneNet1.132.325.433.528.118.721.431.0-0.5930.4940.1400.195/
3DLaneNet1.144.140.846.547.541.532.141.7------
PersFormer1.150.545.658.754.050.041.653.1model0.3190.3250.1120.14189.51
PersFormer1.253.146.858.754.048.441.452.5model0.3610.3280.1240.12988.99
MethodAllUp&<br>DownCurveExtreme<br>WeatherNightIntersectionMerge&<br>Split
LaneATT-S28.325.325.832.027.614.024.3
LaneATT-M31.028.327.434.730.217.026.5
PersFormer42.040.746.343.736.128.941.2
CondLaneNet-S52.355.357.545.846.648.445.5
CondLaneNet-M55.058.559.449.248.650.747.8
CondLaneNet-L59.162.162.954.751.055.752.3
MethodF1(%)Precision(%)Recall(%)CD error(m)Best model
3DLaneNet44.7361.4635.160.127/
GenLaneNet45.5963.9535.420.121/
SALAD (paper of ONCE 3DLanes )64.0775.9055.420.098/
PersFormer72.0777.8267.110.086model

Visualization

Following are the visualization results of PersFormer on OpenLane dataset and Apollo dataset.

Citation

Please use the following citation if you find our repo or our paper PersFormer useful:

    @inproceedings{chen2022persformer,
      title={PersFormer: 3D Lane Detection via Perspective Transformer and the OpenLane Benchmark},
      author={Chen, Li and Sima, Chonghao and Li, Yang and Zheng, Zehan and Xu, Jiajie and Geng, Xiangwei and Li, Hongyang and He, Conghui and Shi, Jianping and Qiao, Yu and Yan, Junchi},
      booktitle={European Conference on Computer Vision (ECCV)},
      year={2022}
    }  

Acknowledgements

We would like to acknowledge the great support from SenseBee labelling team at SenseTime Research, constructive suggestion from Zihan Ding at BUAA, and the fruitful discussions and comments for this project from Zhiqi Li, Yuenan Hou, Yu Liu, Jing Shao, Jifeng Dai. We thank for the code implementation from Gen-LaneNet, LaneATT and Deformable DETR.

License

All code within this repository is under Apache License 2.0.