结合EEG、眼动追踪和高速视频的眼部活动分析数据集
收藏资源简介:
本研究介绍了一个大型的多模态数据集,该数据集结合了脑电图(EEG)信号、眼动追踪、高速相机录像以及参与者的心理状态和特征,以对眼部相关运动进行多因素分析。数据集包含超过46小时的数据,来自31名参与者在63次实验中的数据,每个参与者进行了2520次试验。数据集旨在开发能够有效处理眼部诱导伪影并提高特定任务分类算法的算法。此外,它还提供了评估跨范式鲁棒性的机会,涉及相同的参与者。数据集包括电机想象、电机执行、稳态视觉诱发电位和P300拼写者四种范式。该数据集提供了同时进行电生理记录、视频捕捉和同步眼动追踪的记录,所有数据和代码均可在网上获取以重现实验。
This study introduces a large-scale multimodal dataset that combines electroencephalogram (EEG) signals, eye-tracking, high-speed camera footage, as well as participants’ mental states and characteristics for multi-factor analysis of eye-related movements. The dataset contains over 46 hours of data collected from 31 participants across 63 experimental sessions, with each participant completing 2520 trials. It is designed to develop algorithms that can effectively handle eye-induced artifacts and improve task-specific classification algorithms. Furthermore, it offers the opportunity to evaluate cross-paradigm robustness using the same cohort of participants. The dataset encompasses four paradigms: motor imagery, motor execution, steady-state visual evoked potential (SSVEP), and P300 spellers. It provides simultaneous recordings of electrophysiological signals, video footage, and synchronized eye-tracking data; all datasets and code are publicly available online for experimental reproducibility.

- 1Dataset combining EEG, eye-tracking, and high-speed video for ocular activity analysis across BCI paradigms上海交通大学机械工程系 · 2025年



