Tiny Trap Benchmark

该数据集是一个为评估四足机器人在跑道 上穿越各种小型障碍物或“微陷阱”的能力而设计的基准测试。该基准测试包含三种类型的陷阱：横杆、立杆和坑洞，并采用不同的评价指标，包括成功率、平均通过时间和平均行进距离。该数据集在模拟环境中包含了4,096个四足机器人代理，其任务是在微小陷阱存在的情况下进行机器人运动。

This dataset is a benchmark designed to evaluate the ability of quadruped robots to traverse various small obstacles or "micro-traps" on a runway. This benchmark includes three types of traps: crossbars, vertical poles, and pits, and adopts multiple evaluation metrics including success rate, average passage time, and average travel distance. This dataset contains 4,096 quadruped robot agents in a simulated environment, where the assigned task for these agents is to perform robot locomotion in the presence of such micro-traps.