Kimera-Multi数据集
收藏资源简介:
Kimera-Multi数据集由麻省理工学院航空航天系创建,包含三个大规模多机器人基准数据集,用于评估分布式多机器人视觉SLAM系统。这些数据集记录了在MIT校园进行的现场实验,涉及多达8个机器人,总行程长达8公里,涵盖了多种复杂环境,如地下隧道、室内外混合场景和动态物体。数据集创建过程中,通过精确的参考轨迹和地图进行评估,以确保数据质量。该数据集主要用于研究多机器人系统在现实世界中的可靠部署,特别是在处理间歇性通信、动态环境和硬件故障等挑战时的表现。
The Kimera-Multi dataset was developed by the Department of Aeronautics and Astronautics at the Massachusetts Institute of Technology (MIT). It includes three large-scale multi-robot benchmark datasets intended for evaluating distributed multi-robot visual SLAM systems. These datasets capture field experiments conducted on the MIT campus, involving up to 8 robots with a total traversal distance of 8 kilometers, and cover a diverse array of complex environments such as underground tunnels, mixed indoor-outdoor scenarios, and dynamic objects. To ensure data quality, precise reference trajectories and maps were employed for validation during the dataset development phase. This dataset is primarily utilized for researching the reliable deployment of multi-robot systems in real-world scenarios, especially their performance when addressing challenges including intermittent communication, dynamic environments, and hardware faults.

- 1Resilient and Distributed Multi-Robot Visual SLAM: Datasets, Experiments, and Lessons Learned麻省理工学院航空航天系 · 2023年



