自动驾驶多模态多传感器标注数据集

本数据集是面向城市自动驾驶的多传感器标注数据集，覆盖昼夜、晴雨等复杂路况场景，包含相机与激光雷达同步采集数据。完成 2D/3D 障碍物、车道线、可行驶区域等全要素标注，含多类交通参与者目标，经过多级质检保障标注质量。数据划分训练 / 验证 / 测试子集，格式标准化，适用于自动驾驶感知、融合、预测模型训练与算法评测。This dataset is a multi-sensor annotated dataset for urban autonomous driving, covering complex road scenarios such as day and night, sunny and rainy weather. It includes synchronously collected data from cameras and LiDAR. It provides comprehensive annotations of various elements including 2D/3D obstacles, lane lines and drivable areas, covering multiple types of traffic participant targets. The annotation quality is guaranteed through multi-level quality inspections. The data is divided into training, validation and test subsets with standardized formats, which is applicable for the training and algorithm evaluation of autonomous driving perception, fusion and prediction models.

搜集汇总

背景与挑战

背景概述

该数据集是针对城市自动驾驶场景的多传感器标注数据集，同步采集相机与激光雷达数据，覆盖昼夜、晴雨等多种复杂路况。它提供了2D/3D障碍物、车道线、可行驶区域等全要素标注，并经过多级质量检查，数据划分为训练、验证和测试子集，适用于自动驾驶感知、融合及预测模型的训练与评估。