BezierFormer: A Unified Architecture for 2D and 3D Lane Detection

TL;DR

BezierFormer introduces a unified architecture for 2D and 3D lane detection using Bézier curve representation, novel attention mechanisms, and a Chamfer IoU loss, achieving state-of-the-art results.

Contribution

It is the first unified model for both 2D and 3D lane detection leveraging Bézier curves, with innovative attention and loss functions.

Findings

Achieves state-of-the-art performance on 2D and 3D lane detection benchmarks.

Demonstrates the effectiveness of Bézier curve representation for lane modeling.

Validates the proposed Chamfer IoU loss for better control point regression.

Abstract

Lane detection has made significant progress in recent years, but there is not a unified architecture for its two sub-tasks: 2D lane detection and 3D lane detection. To fill this gap, we introduce B\'{e}zierFormer, a unified 2D and 3D lane detection architecture based on B\'{e}zier curve lane representation. B\'{e}zierFormer formulate queries as B\'{e}zier control points and incorporate a novel B\'{e}zier curve attention mechanism. This attention mechanism enables comprehensive and accurate feature extraction for slender lane curves via sampling and fusing multiple reference points on each curve. In addition, we propose a novel Chamfer IoU-based loss which is more suitable for the B\'{e}zier control points regression. The state-of-the-art performance of B\'{e}zierFormer on widely-used 2D and 3D lane detection benchmarks verifies its effectiveness and suggests the worthiness of further exploration.