Hierarchical Particle-Based Environment

该数据集是通过一个基于FleX物理引擎的自定义交互式粒子环境生成的训练数据组成的，它能够实现对各种物体和环境属性下物理动力学的预测。此外，该数据集还包含了刚体运动学、复杂形状、动态碰撞、多对象交互、软体、布料以及流体的示例。该任务旨在预测模拟环境中物体的物理动力学。

This dataset comprises training data generated using a custom interactive particle environment based on the FleX physics engine, which supports physical dynamics prediction across various object and environmental properties. Furthermore, this dataset includes examples of rigid body kinematics, complex geometries, dynamic collisions, multi-object interactions, soft bodies, cloth, and fluids. The task aims to predict the physical dynamics of objects within simulated environments.