Simulated Trajectories Dataset

该数据集包含了在2D空间中虚拟移动代理的可能轨迹，以及一个模拟的8自由度机械臂的数据。这些数据被用于训练和测试目标导向的规划模型。数据集还包括了手工绘制的轨迹，这些轨迹被重新采样至30个时间步长，旨在评估生成模型在均方根误差和目标偏差方面的表现。数据集规模涵盖多条轨迹，这些轨迹具有不同的噪声水平和目标点分布。任务旨在进行目标导向的轨迹生成及评估。

This dataset encompasses possible trajectories of virtual mobile agents operating in 2D space, alongside data from a simulated 8-degree-of-freedom (8-DOF) robotic arm. These data are utilized for training and testing goal-oriented planning models. The dataset also includes hand-drawn trajectories that have been resampled to 30 timesteps, aimed at evaluating the performance of generative models in terms of root mean squared error (RMSE) and target deviation. The dataset comprises multiple trajectories with distinct noise levels and target point distributions. The task is designed for goal-oriented trajectory generation and evaluation.