SynDrone
收藏资源简介:
SynDrone是一个专为城市场景设计的无人机多模态数据集,由帕多瓦大学信息工程系创建。该数据集包含72000个标注样本,涵盖多种飞行高度和视角,支持深度学习模型的有效训练。数据集内容包括高分辨率图像和3D数据,具有28个类别的像素级标注,适用于语义分割等任务。创建过程中使用了CARLA模拟器,确保数据的真实性和多样性。SynDrone的应用领域广泛,包括城市规划、交通监控和环境监测等,旨在解决无人机在复杂城市环境中的视觉理解问题。
SynDrone is a multimodal drone dataset specifically designed for urban scenarios, created by the Department of Information Engineering at the University of Padua. This dataset comprises 72,000 annotated samples covering various flight altitudes and viewing perspectives, which supports effective training of deep learning models. The dataset includes high-resolution images and 3D data, with pixel-level annotations across 28 categories, making it suitable for tasks such as semantic segmentation. The CARLA simulator was utilized during its development to ensure the authenticity and diversity of the data. SynDrone has a wide range of application scenarios including urban planning, traffic monitoring, environmental monitoring and more, aiming to address the visual understanding challenges of drones in complex urban environments.

- 1SynDrone -- Multi-modal UAV Dataset for Urban Scenarios帕多瓦大学信息工程系 · 2023年



