无人驾驶算法模型训练路侧数据

通过自动驾驶示范区道路的路侧设备采集的场景数据，经过抽帧后的图片，用描述文件标注出其中的要素，如：小汽车、SUV、大巴车、摩托车、骑手、行人等要素，可用于无人驾驶算法模型训练，提高算法精准度。

Scene data collected by roadside equipment on roads within autonomous driving demonstration zones consists of images obtained via frame extraction. These images are annotated using descriptive files to mark various elements including cars, SUVs, buses, motorcycles, riders, pedestrians and other relevant objects. This dataset can be utilized for training autonomous driving algorithm models to enhance their accuracy.

搜集汇总

背景与挑战

背景概述

该数据集由上海金桥智能网联汽车发展有限公司提供，包含通过自动驾驶示范区路侧设备采集的场景数据，经抽帧和标注后形成图片及描述文件，覆盖小汽车、SUV、大巴车、摩托车、骑手、行人等要素，适用于无人驾驶算法模型训练以提升精准度。数据覆盖上海金桥区域部分路段，存储大小为300 MB，更新频率为静态。