GroupLane: End-to-End 3D Lane Detection with Channel-wise Grouping

TL;DR

GroupLane is a fast, end-to-end 3D lane detection method that uses row-wise classification in bird-eye-view, supporting both vertical and horizontal lanes with high accuracy and efficiency.

Contribution

It introduces a fully convolutional, end-to-end 3D lane detection framework with a novel row-wise classification approach and lane-aware grouping, outperforming state-of-the-art methods.

Findings

Outperforms PersFormer by 13.6% F1 score on OpenLane.

Achieves nearly 7x faster inference with significantly fewer FLOPs.

Supports detection of both vertical and horizontal lanes in bird-eye-view.

Abstract

Efficiency is quite important for 3D lane detection due to practical deployment demand. In this work, we propose a simple, fast, and end-to-end detector that still maintains high detection precision. Specifically, we devise a set of fully convolutional heads based on row-wise classification. In contrast to previous counterparts, ours supports recognizing both vertical and horizontal lanes. Besides, our method is the first one to perform row-wise classification in bird-eye-view. In the heads, we split feature into multiple groups and every group of feature corresponds to a lane instance. During training, the predictions are associated with lane labels using the proposed single-win one-to-one matching to compute loss, and no post-processing operation is demanded for inference. In this way, our proposed fully convolutional detector, GroupLane, realizes end-to-end detection like DETR. Evaluated on 3 real world 3D lane 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 nearly 7x faster and the FLOPs is only 13.3% of it.