Grouplane: End-to-End 3D Lane Detection with Channel-Wise Grouping

TL;DR — This paper develops a new way of modeling 3D lanes and proposes a new 3D lane detector, which achieves SOTA performance.

Abstract

Efficiency is quite important for 3D lane detection while previous detectors are either computationally expensive or difficult for optimization. To bridge this gap, we propose a fully convolutional detector named GroupLane, which is simple, fast, and still maintains high detection precision. Specifically, we first propose to split extracted feature into multiple groups along the channel dimension and employ every group to represent a prediction. In this way, GroupLane realizes end-to-end detection like DETR based on pure convolutional neural network. Then, we propose to represent lanes by performing row-wise classification in bird’s eye view and devise a set of corresponding detection heads. Compared with existing row-wise classification implementations that only support recognizing vertical lanes, ours can detect both vertical and horizontal ones. Additionally, a matching algorithm named single-win one-to-one matching is developed to associate prediction with labels during training. Evaluated on 3 benchmarks, OpenLane, Once-3DLanes, and OpenLane-Huawei, GroupLane adopting ConvNext-Base as the backbone outperforms the published state-of-the-art PersFormer by 13.6% F1 score in the OpenLane validation set. Besides, GroupLane with ResNet18 still surpasses PersFormer by 4.9% F1 score, while the inference speed is 7$\times$ faster.