V2X-Radar
收藏资源简介:
V2X-Radar是由清华大学车辆与运载学院创建的第一个大规模真实世界多模态数据集,专注于合作感知中的4D雷达数据。该数据集包含20K LiDAR帧、40K摄像头图像和20K 4D雷达数据,涵盖了晴天、雨天等多种天气条件和白天、黄昏、夜晚等不同时间段。数据集通过连接车辆平台和智能路边单元收集,经过精心筛选和标注,包含350K个标注的3D边界框,涵盖五类目标。V2X-Radar数据集分为三个子集:V2X-Radar-C用于合作感知,V2X-Radar-I用于路边感知,V2X-Radar-V用于单车辆感知,旨在解决自动驾驶中的遮挡和感知范围有限的问题。
V2X-Radar is the first large-scale real-world multimodal dataset created by the School of Vehicle and Mobility, Tsinghua University, focusing on 4D radar data in cooperative perception. This dataset contains 20K LiDAR frames, 40K camera images and 20K 4D radar data, covering various weather conditions such as sunny and rainy days, as well as different time periods including daytime, dusk and nighttime. Collected via connected vehicle platforms and intelligent roadside units, the dataset has been carefully screened and annotated, containing 350K annotated 3D bounding boxes covering five types of targets. The V2X-Radar dataset is divided into three subsets: V2X-Radar-C for cooperative perception, V2X-Radar-I for roadside perception, and V2X-Radar-V for single-vehicle perception, aiming to address the issues of occlusion and limited perception range in autonomous driving.

- 1V2X-Radar: A Multi-modal Dataset with 4D Radar for Cooperative Perception清华大学车辆与运载学院 · 2024年



