WEMAC

WEMAC数据集是由卡洛斯三世大学性别研究学院创建的，专注于女性情感的多模态情感计算数据集。该数据集包含100名女性的生理、语音和自我报告的情感数据记录，旨在检测基于性别的暴力。数据集通过使用沉浸式虚拟现实头盔展示经过验证的视频集来诱发真实情感。生理信号包括血容量脉搏、皮肤电反应和皮肤温度。语音数据在刺激可视化后立即获取，以捕捉感知情感的最终痕迹。参与者被要求在12种分类情感中进行标注，使用修改版的自我评估小人图进行几种维度情感标注，以及喜欢和熟悉度标签。数据集的应用领域包括情感计算、基于性别的暴力解决方案设计、主观和自我报告情感标注的理解以及女性恐惧分类。

The WEMAC dataset, a multimodal affective computing dataset focused on women’s emotions, was developed by the Institute for Gender Studies of Charles III University of Madrid. It contains physiological, vocal, and self-reported emotional data records from 100 female participants, with the primary objective of detecting gender-based violence. To elicit authentic emotional states, the dataset presents a validated set of video stimuli via immersive virtual reality headsets. Collected physiological signals include blood volume pulse, galvanic skin response, and skin temperature. Vocal data is collected immediately after the visual stimulation to capture the final traces of the participants’ perceived emotions. Participants were required to complete three annotation tasks: labeling 12 categorical emotions, conducting dimensional emotion assessments using a modified version of the Self-Assessment Manikin (SAM), and rating their perceived liking and familiarity toward the stimuli. The applicable research domains of this dataset include affective computing, the design of gender-based violence intervention solutions, the investigation of subjective and self-reported emotion annotation methodologies, and the classification of women’s fear responses.

搜集汇总

背景与挑战

背景概述

WEMAC是一个专注于女性情感的多模态情感计算数据集，由卡洛斯三世大学性别研究学院创建，包含100名女性的生理、语音和自我报告情感数据，旨在通过虚拟现实诱发真实情感以检测基于性别的暴力。数据集涵盖血容量脉搏等生理信号、语音数据以及多种情感标注，应用于情感计算和暴力解决方案设计等领域。