Facial Expressions in Real News Dataset

本研究介绍了名为“Facial Expressions in Real News Dataset”的数据集，由澳大利亚国立大学创建，旨在通过分析新闻视频中的面部表情来提高情感识别的准确性。该数据集包含14,172个实例，涵盖了从国际知名媒体下载的318个新闻视频中提取的面部图像。数据集通过严格的质量控制和面部特征提取，确保了数据的高质量和一致性。该数据集主要用于情感识别技术的研究和开发，特别是在智能设备和实时交互系统中的应用。

This study introduces a dataset named "Facial Expressions in Real News Dataset", which was developed by The Australian National University. The dataset is designed to improve the accuracy of emotion recognition by analyzing facial expressions in news videos. It contains 14,172 instances, consisting of facial images extracted from 318 news videos downloaded from internationally recognized media outlets. Strict quality control and facial feature extraction workflows are implemented throughout the dataset construction process to ensure its high quality and consistency. This dataset is mainly utilized for the research and development of emotion recognition technologies, especially for applications in smart devices and real-time interactive systems.