3D_Lane_Synthetic_Dataset

Dataset Overview

Dataset Name

A Synthetic Dataset for 3D Lane Detection

Dataset Purpose

To promote the development and evaluation of 3D lane detection methods.

Dataset Source

Extended from the Apollo Synthetic Dataset, download links include Google Drive and Baidu Netdisk.

Construction & Evaluation Method

Based on the ECCV 2020 paper “Gen‑LaneNet: a generalized and scalable approach for 3D lane detection”.

Dataset Structure

• Data Splits: Standard, rare subsets, and illumination‑variation splits are provided.
• Data Preparation: Includes parsing raw data, preparing splits, and preparing subsets.

Evaluation Method

• Evaluation Script: eval_3D_lane.py
• Metrics: Average Precision (AP), maximum F‑score, x and z errors (near and far).
• Results: Compare two baseline methods across three split types.

Visualization

Supports visual comparison with ground truth by setting vis = True.

Citation

@article{guo2020gen,
  title={Gen‑LaneNet: A Generalized and Scalable Approach for 3D Lane Detection},
  author={Yuliang Guo, Guang Chen, Peitao Zhao, Weide Zhang, Jinghao Miao, Jinghao Wang, and Tae Eun Choe},
  booktitle={Computer Vision – ECCV 2020 – 16th European Conference},
  year={2020}
}