The 1st-place Solution for CVPR 2023 OpenLane Topology in Autonomous Driving Challenge

TL;DR

This paper presents a top-performing solution for autonomous driving topology prediction, combining centerline detection with traffic element detection using advanced detectors and a novel MLP-based topology predictor, achieving state-of-the-art results.

Contribution

The paper introduces a multi-stage framework integrating PETRv2, YOLOv8, and an MLP head for improved topology reasoning in autonomous driving.

Findings

Achieved 55% OLS on OpenLaneV2 test set.

Surpassed second-place solution by 8 points.

Demonstrated effectiveness of combined detection and topology prediction.

Abstract

We present the 1st-place solution of OpenLane Topology in Autonomous Driving Challenge. Considering that topology reasoning is based on centerline detection and traffic element detection, we develop a multi-stage framework for high performance. Specifically, the centerline is detected by the powerful PETRv2 detector and the popular YOLOv8 is employed to detect the traffic elements. Further, we design a simple yet effective MLP-based head for topology prediction. Our method achieves 55\% OLS on the OpenLaneV2 test set, surpassing the 2nd solution by 8 points.