Quadruped Robot Skill Learning Dataset
收藏arXiv2025-09-30 收录
下载链接：
https://robot-skills.github.io
下载链接
链接失效反馈官方服务：
资源简介：
该数据集包含了四足机器人在不同技能学习方法下，在模拟和真实世界环境中性能的定量分析。其中涵盖了多种指标，如速度跟踪误差、轨迹跟踪误差以及在不同技能执行条件下的回归误差。这些数据旨在评估在运动和操作任务中技能执行的效果。
This dataset provides quantitative analyses of the performance of quadruped robots employing different skill learning methods in both simulated and real-world environments. It includes a range of metrics such as velocity tracking error, trajectory tracking error, and regression error under diverse skill execution conditions. This dataset is intended to assess the efficacy of skill execution during locomotion and manipulation tasks.
提供机构：
Unitree Robotics and IsaacGym
