Anti-UAV
收藏资源简介:
Anti-UAV数据集由中国科学院大学创建,专注于无人机跟踪研究。该数据集包含318对RGB-T视频,总计超过580,000个手动标注的边界框,适用于长距离无人机跟踪。数据集内容丰富,包括多种场景和光照条件下的视频序列,支持单模态和多模态无人机跟踪。创建过程中,数据集通过精细的标注策略确保高质量。该数据集的应用领域主要集中在无人机监控和跟踪技术的发展,旨在解决无人机在复杂环境中的状态感知问题。
The Anti-UAV dataset, developed by the University of Chinese Academy of Sciences (UCAS), focuses on unmanned aerial vehicle (UAV) tracking research. This dataset comprises 318 pairs of RGB-T video sequences, with a total of over 580,000 manually annotated bounding boxes, and is tailored for long-distance UAV tracking. It encompasses diverse content, including video sequences under various scenarios and lighting conditions, supporting both unimodal and multimodal UAV tracking tasks. During its development, a rigorous annotation strategy was employed to ensure high-quality annotations. The primary application scope of this dataset centers on the advancement of UAV monitoring and tracking technologies, aiming to address the challenge of UAV situational awareness in complex environments.

- 1Anti-UAV: A Large Multi-Modal Benchmark for UAV Tracking中国科学院大学 · 2021年



