SMIC-VIS-E dataset
收藏资源简介:
本研究构建的SMIC-VIS-E数据集由芬兰奥卢大学信息技术和电气工程学院创建,包含76个视频序列,主要用于微表情检测研究。数据集中的视频序列大小为640×480像素,记录帧率为25fps,其中71个视频包含微表情,5个为非微表情视频。这些视频均已标注起始和结束帧,适用于面部表情分析和机器学习模型的训练与评估。数据集的创建旨在标准化微表情检测方法的性能评估,推动相关技术的发展。
The SMIC-VIS-E dataset constructed in this study was created by the School of Information Technology and Electrical Engineering, University of Oulu, Finland. It consists of 76 video sequences, primarily intended for microexpression detection research. The video sequences in the dataset have a resolution of 640×480 pixels and a recording frame rate of 25 fps. Among them, 71 videos contain microexpressions, while the remaining 5 are non-microexpression videos. All these videos have been annotated with their start and end frames, making them suitable for facial expression analysis as well as the training and evaluation of machine learning models. The creation of this dataset aims to standardize the performance evaluation of microexpression detection methods and promote the development of related technologies.

- 1Micro-Expression Spotting: A Benchmark信息技术和电气工程学院 · 2017年



