Enhancing 3D Lane Detection and Topology Reasoning with 2D Lane Priors

TL;DR

Topo2D introduces a Transformer-based framework that leverages 2D lane priors to improve 3D lane detection and topology reasoning, significantly outperforming existing methods on key benchmarks.

Contribution

The paper presents Topo2D, a novel approach that incorporates 2D lane priors into 3D lane detection and topology reasoning using Transformer architecture.

Findings

Achieves 44.5% OLS on OpenLane-V2 topology benchmark

Attains 62.6% F-score on OpenLane 3D lane detection

Outperforms previous state-of-the-art methods

Abstract

3D lane detection and topology reasoning are essential tasks in autonomous driving scenarios, requiring not only detecting the accurate 3D coordinates on lane lines, but also reasoning the relationship between lanes and traffic elements. Current vision-based methods, whether explicitly constructing BEV features or not, all establish the lane anchors/queries in 3D space while ignoring the 2D lane priors. In this study, we propose Topo2D, a novel framework based on Transformer, leveraging 2D lane instances to initialize 3D queries and 3D positional embeddings. Furthermore, we explicitly incorporate 2D lane features into the recognition of topology relationships among lane centerlines and between lane centerlines and traffic elements. Topo2D achieves 44.5% OLS on multi-view topology reasoning benchmark OpenLane-V2 and 62.6% F-Socre on single-view 3D lane detection benchmark OpenLane, exceeding the performance of existing state-of-the-art methods.