V2X-Seq
收藏资源简介:
V2X-Seq是由清华大学智能产业研究院推出的首个大规模连续V2X数据集,专注于车辆与基础设施协同感知与预测。该数据集包含超过15,000帧的数据,涵盖了95个场景,包括基础设施和车辆侧的图像、点云、3D检测/跟踪标注及矢量地图。此外,还有约210,000个场景用于轨迹预测,这些数据来自28个交叉口区域,总计672小时的数据。V2X-Seq旨在通过车辆与基础设施的协同,解决自动驾驶中的感知和预测问题,为车辆-基础设施协同自动驾驶社区提供了一个理想的研究和测试平台。
V2X-Seq is the first large-scale sequential V2X dataset developed by the Institute for AI Industry Research (AIR) of Tsinghua University, focusing on vehicle-infrastructure cooperative perception and prediction. This dataset comprises over 15,000 frames of data across 95 scenarios, including images, point clouds, 3D detection and tracking annotations, as well as vector maps from both infrastructure and vehicle sides. In addition, approximately 210,000 scenarios for trajectory prediction are collected from 28 intersection areas, totaling 672 hours of data. V2X-Seq aims to solve perception and prediction challenges in autonomous driving through vehicle-infrastructure collaboration, providing an ideal research and testing platform for the vehicle-infrastructure cooperative autonomous driving community.

- 1V2X-Seq: A Large-Scale Sequential Dataset for Vehicle-Infrastructure Cooperative Perception and Forecasting清华大学人工智能产业研究院 · 2023年



