EEG data for observing the video stimuli

This is EEG dataset collected within the research "Natural image reconstruction from brain waves: a novel visual BCI system with native feedback"
https://www.biorxiv.org/content/10.1101/787101v3
Here we propose that observing visual stimuli of 5 different categories results in the different brain wave patterns decodable from noninvasive EEG. This hypothesis was tested on 17 subjects.
All the data was collected by in autumn 2018, at MIPT Neurorobotics Lab, Dolgoprudny, Moscow Region, Russian Federation.
The data was acquired using 128-channel MCS EEG Cap and NVX136 MCS amplifier, using NeoRec software provided by the manufacturer.
The protocol included two sessions of the same video observing task. The video consisted of 117 different short clips, belonging to 5 different categories: abstract forms, waterfalls, faces, Goldberg mechanisms and speed. The clips were separated with black screen pauses of 2-3 seconds length. As for now, we cannot include the video itself due to copyright issues. We only provide screenshots of the video clips (except for category 2: faces).

The detailed description can be found inside the ZIP archive.

本数据集系于“自然图像重建自脑电波：一种带有原生反馈功能的新型视觉脑机接口系统”研究项目中收集的脑电图（EEG）数据集。该研究旨在提出并验证一个假设，即观察五种不同类别的视觉刺激将导致可从非侵入性脑电图解码出的不同脑电波模式。该假设在17名受试者身上进行了测试。所有数据均于2018年秋季在俄罗斯联邦莫斯科地区多尔戈普鲁德尼的莫斯科物理技术学院（MIPT）神经机器人实验室采集。数据采集采用128通道MCS脑电图帽和NVX136 MCS放大器，使用由制造商提供的NeoRec软件进行。实验方案包括两次相同视频观察任务。视频由117个不同短片剪辑组成，分为五大类别：抽象形态、瀑布、人脸、戈德堡机构和速度。剪辑之间以2-3秒的黑屏暂停分隔。由于版权问题，目前无法包含视频本身，仅提供视频剪辑的截图（第二类：人脸除外）。详细描述可于ZIP压缩文件中查阅。