CG-MER
收藏资源简介:
CG-MER数据集是一个基于卡牌游戏的多模态情感识别数据集,由法国CESI LINEACT实验室创建。该数据集包含20名参与者(9名女性,11名男性)在10次会话中生成的面部表情、语音和手势数据,总时长约10小时。数据通过卡牌游戏收集,参与者被要求表达多种情感并回答不同强度的问题。数据集涵盖了RGB视频、深度视频和音频文件,并包含自我、伙伴和外部观察者的情感注释。该数据集旨在解决情感识别中的多模态数据融合问题,适用于情感计算、人机交互和心理健康等领域的研究。
The CG-MER dataset is a multi-modal emotion recognition dataset based on card games, created by the CESI LINEACT laboratory in France. The dataset includes facial expression, voice, and gesture data generated by 20 participants (9 females and 11 males) across 10 sessions, with a total duration of approximately 10 hours. The data was collected through card games, where participants were tasked with expressing a variety of emotions and answering questions of different intensities. The dataset encompasses RGB video, depth video, and audio files, and includes emotional annotations from self, partner, and external observers. The dataset aims to address the issue of multi-modal data fusion in emotion recognition and is suitable for research in areas such as affective computing, human-computer interaction, and mental health.




