DOLPHINS
收藏资源简介:
DOLPHINS数据集是由清华大学电子工程系创建的一个大规模、多场景、多视角、多模态的自动驾驶数据集。该数据集包含42376帧图像和点云数据,涵盖6种典型自动驾驶场景,如城市交叉口、T型路口等,并考虑了动态天气条件。数据集通过CARLA模拟器生成,确保了数据的多样性和真实性。DOLPHINS数据集旨在支持车辆间(V2V)和车辆与基础设施间(V2I)的协同感知研究,解决自动驾驶中的盲区和长距离感知问题,推动互联自动驾驶技术的发展。
The DOLPHINS dataset is a large-scale, multi-scenario, multi-view, multi-modal autonomous driving dataset created by the Department of Electronic Engineering of Tsinghua University. It contains 42,376 frames of image and point cloud data, covering 6 typical autonomous driving scenarios such as urban intersections and T-junctions, and takes into account dynamic weather conditions. The dataset is generated via the CARLA simulator, ensuring its diversity and authenticity. The DOLPHINS dataset aims to support research on cooperative perception between vehicles (V2V) and between vehicles and infrastructure (V2I), addressing the issues of blind spots and long-distance perception in autonomous driving, and promoting the development of connected autonomous driving technologies.

- 1DOLPHINS: Dataset for Collaborative Perception enabled Harmonious and Interconnected Self-driving清华大学电子工程系 · 2022年



