MFA-Wild
收藏arXiv2021-12-10 更新2024-08-06 收录
下载链接：
http://arxiv.org/abs/2112.05267v1
下载链接
链接失效反馈官方服务：
资源简介：
MFA-Wild是由西蒙弗雷泽大学计算机科学学院创建的第一个野外多元文化视频情感数据集，专注于愤怒相关情感表达。该数据集包含198个来自波斯和北美文化的视频，每个视频时长1至10秒。数据集的创建过程涉及从YouTube收集视频，并通过设计网络界面进行文化流利的标注。MFA-Wild旨在通过多标签框架深入探索情感表达，解决情感识别任务中对多样性和文化差异的理解问题。
MFA-Wild is the first in-the-wild multicultural video emotion dataset focused on anger-related affective expressions, created by the School of Computer Science at Simon Fraser University. This dataset comprises 198 videos sourced from Persian and North American cultural backgrounds, with each clip lasting between 1 and 10 seconds. The dataset construction process involves collecting videos from YouTube and performing culturally fluent annotations via a custom-designed web interface. MFA-Wild aims to conduct in-depth investigations into emotional expressions through a multi-label framework, addressing the gap in understanding diversity and cultural differences in emotion recognition tasks.
提供机构：
西蒙弗雷泽大学计算机科学学院
创建时间：
2021-12-10
