PathBench
收藏资源简介:
PathBench是由帝国理工学院和Ryerson大学合作开发的一个用于评估和比较经典及学习型路径规划算法的平台。该数据集支持2D和3D环境下的算法开发、可视化、训练和测试,并提供对机器人操作系统(ROS)的支持。PathBench包含多种内置地图和外部环境,如视频游戏和现实世界数据库,用于评估算法的性能指标,如路径长度、成功率、计算时间和路径偏差。此外,PathBench还提供了一个ROS实时扩展,以便与真实世界的机器人进行交互,旨在解决路径规划领域的算法比较和性能评估问题。
PathBench is a platform jointly developed by Imperial College London and Ryerson University for evaluating and comparing both classical and learning-based path planning algorithms. This platform supports algorithm development, visualization, training and testing in 2D and 3D environments, and provides support for the Robot Operating System (ROS). PathBench includes a variety of built-in maps and external environments, such as video games and real-world databases, for evaluating algorithm performance metrics including path length, success rate, computation time and path deviation. Furthermore, PathBench provides a real-time ROS extension to enable interaction with real-world robots, aiming to address the challenges of algorithm comparison and performance evaluation in the path planning domain.

- 1PathBench: A Benchmarking Platform for Classical and Learned Path Planning Algorithms帝国理工学院 · 2021年



