Esports Trajectories and Actions (ESTA)
收藏arXiv2022-09-21 更新2024-06-21 收录
下载链接：
https://github.com/pnxenopoulos/esta
下载链接
链接失效反馈官方服务：
资源简介：
ESTA数据集是由纽约大学开发的，专注于电子竞技中的轨迹和动作数据。该数据集包含从1558场专业Counter-Strike比赛中提取的8.6百万次玩家动作、7.9百万个游戏帧和41.7万个轨迹。数据通过游戏服务器日志获取，具有高度的清洁度和详细性。ESTA数据集适用于多种应用，如玩家特定信息的胜利预测，为电子竞技分析提供了一个大规模且详细的公共资源。
The ESTA dataset, developed by New York University, focuses on trajectory and action data in esports. It contains 8.6 million player actions, 7.9 million game frames, and 417,000 trajectories extracted from 1,558 professional Counter-Strike matches. The data is collected via game server logs, featuring high cleanliness and detailed granularity. The ESTA dataset supports various applications such as victory prediction based on player-specific information, serving as a large-scale and detailed public resource for esports analytics.
提供机构：
纽约大学
创建时间：
2022-09-21
