Dribbling Policies Dataset

该数据集用于评估在具有不同球阻系数和地形的模拟环境中运球策略的效果。它允许在不同模拟条件下，如不同地形和阻系数下，比较不同算法在运球任务中的表现。该数据集规模宏大，包含8,000个并行训练的领域随机化智能体，其任务专注于动态物体操作，在足球运球场景中尤为关键。

This dataset is developed to evaluate the effectiveness of dribbling strategies in simulated environments featuring varying ball resistance coefficients and terrain types. It facilitates comparative performance analysis of diverse algorithms for dribbling tasks across a range of simulated conditions, including different terrains and ball resistance coefficients. Boasting a large scale, this dataset encompasses 8,000 parallel-trained domain-randomized AI Agents, whose tasks center on dynamic object manipulation—a capability that is particularly critical in soccer dribbling scenarios.