Robot Soccer Simulation Dataset
收藏arXiv2025-09-30 收录
官方服务：
资源简介：
该数据集是通过模拟生成的，旨在训练基于视觉的智能体利用深度强化学习来踢足球。该数据集包含了源自第一人称视角的复杂行为，使得可以发现更复杂的足球策略。这是一份大规模的模拟数据，其任务是对多代理机器人足球进行训练和评估。
This dataset is synthetically generated for training vision-based AI Agents to play soccer via deep reinforcement learning. It encompasses complex first-person perspective behaviors, facilitating the discovery of more advanced soccer strategies. As a large-scale simulated dataset, it is dedicated to the training and evaluation of multi-agent robotic soccer systems.
