边缘计算单元功能和性能测试数据集

数据集为聚焦于智能网联汽车路侧边缘计算单元感知功能与性能的真实道路测试数据集。数据采集于2023年1月至12月，覆盖某高速公路连续12公里路段，通过部署于路侧的毫米波雷达与摄像头等多模态传感器，同步采集交通参与者信息及高精度真值数据。数据集包含3个典型时段的边缘感知原始输出与对应真值评价数据，旨在为边缘计算单元在目标检测、跟踪及状态估计等方面的感知精度、稳定性与实时性提供多维度评估基准。数据经严格时空同步与质量控制，适用于智能网联汽车路侧感知算法研发、边缘计算单元性能评估与车路协同系统标准化测试。This dataset is a real-world road testing dataset focusing on the perception functions and performance of roadside edge computing units for connected and automated vehicles (CAVs). The data was collected from January to December 2023, covering a continuous 12-kilometer section of a highway. Multi-modal sensors such as millimeter-wave radars and cameras deployed on the roadside are used to synchronously collect information of traffic participants and high-precision ground truth data. The dataset contains raw outputs of edge perception and corresponding ground truth evaluation data from three typical time periods. It aims to provide a multi-dimensional evaluation benchmark for the perception accuracy, stability and real-time performance of edge computing units in tasks including object detection, tracking and state estimation. The data has undergone strict spatiotemporal synchronization and quality control, and is applicable to the research and development of roadside perception algorithms for connected and automated vehicles, performance evaluation of edge computing units, and standardized testing of vehicle-road cooperative systems.

搜集汇总

背景与挑战

背景概述

该数据集是针对智能网联汽车路侧边缘计算单元感知功能与性能的真实道路测试数据集，采集于2023年某高速公路连续12公里路段，通过多模态传感器同步获取交通参与者信息和高精度真值数据。它包含3个典型时段的边缘感知原始输出与真值评价，旨在为目标检测、跟踪及状态估计等任务提供多维度评估基准，适用于算法研发、性能评估和车路协同系统标准化测试。