eSports Sensors Dataset

eSports Sensors Dataset是由斯科尔科沃科学技术研究所和德国人工智能研究中心共同创建的多模态数据集，专注于电子竞技领域。该数据集包含来自22场《英雄联盟》游戏的专业和业余玩家的心理生理数据，总记录时长超过40小时。数据收集包括玩家的生理活动（如运动、脉搏、扫视）、自我报告的赛后调查以及游戏内数据。数据集的一个重要特点是能够同时从五名玩家收集数据，便于团队层面的传感器数据分析。该数据集的应用领域包括技能预测、玩家再识别和团队动态分析，旨在解决电子竞技中的训练和分析问题。

eSports Sensors Dataset is a multimodal dataset co-created by the Skolkovo Institute of Science and Technology and the German Research Center for Artificial Intelligence (DFKI), focusing on the esports domain. This dataset contains psychophysiological data from both professional and amateur players across 22 League of Legends matches, with a total recording duration exceeding 40 hours. The collected data includes players' physiological activities such as movement, pulse, and eye saccades, self-reported post-match surveys, as well as in-game data. An important feature of this dataset is that it enables simultaneous data collection from five players per team, facilitating sensor data analysis at the team level. Its application areas include skill prediction, player re-identification, and team dynamics analysis, aiming to address training and analysis issues in esports.