LLaMAC: Low-cost Biosignal Sensor based Large Multimodal Dataset for Affective Computing

This dataset consists of biosignals such as electroencephalogram (EEG), electrodermal activity (EDA), photoplethysmogram (PPG), skin temperature (SKT), and respiration (RESP), and questionnaire results on continuous domain emotion (valence, arousal, and dominance), discrete domain emotion (type and intensity of emotion: neutral, fun, sadness, angry, and fear), liking, and familiarity. This dataset has the characteristics of using low-cost biosignal sensors, conducting questionnaires on continuous and discrete domain emotions, participating in more than 100 participants, and conducting questionnaires on liking and familiarity. This dataset can be used to predict continuous and discrete domain emotions based on biosignals, analyze correlations between continuous and discrete domain emotions, predict liking based on biosignals, and identify differences in biosignals according to familiarity and predict emotions or liking using them.

本数据集涵盖脑电图（EEG）、皮肤电活动（EDA）、光电容积脉搏波（PPG）、皮肤温度（SKT）与呼吸（RESP）等多类生物信号，同时包含针对连续维度情绪（效价、唤醒度与支配度）、离散维度情绪（情绪类型与强度：中性、愉悦、悲伤、愤怒与恐惧）、喜好度及熟悉度的问卷调研结果。该数据集具有如下特征：采用低成本生物信号传感器采集数据，针对连续与离散维度情绪设计问卷调研，招募超100名受试者参与，并设置了针对喜好度与熟悉度的问卷条目。本数据集可用于以下研究场景：基于生物信号预测连续与离散维度情绪、分析连续与离散维度情绪间的相关性、基于生物信号预测喜好度，以及依据熟悉度区分生物信号差异并借此实现情绪或喜好度的预测。