SAMM Long Videos
收藏资源简介:
SAMM Long Videos数据集是由曼彻斯特都会大学计算机与数学系创建,专注于自发面部微观和宏观表情的识别与检测。该数据集包含147个长视频,总计343个宏观表情和159个微观表情,采用FACS编码详细动作单元。数据集创建过程中,使用OpenFace工具进行面部对齐和动作单元检测,提高了数据处理的质量。该数据集主要应用于情感研究、行为心理学、安全及通信等领域,旨在解决面部微观和宏观表情识别的挑战。
The SAMM Long Videos Dataset was developed by the Department of Computer Science and Mathematics, Manchester Metropolitan University, focusing on the recognition and detection of spontaneous facial micro-expressions and macro-expressions. This dataset includes 147 long videos, comprising a total of 343 macro-expressions and 159 micro-expressions, with detailed Action Units (AUs) annotated using the Facial Action Coding System (FACS). During the dataset construction process, the OpenFace tool was employed for facial alignment and Action Unit detection, which improved the quality of data processing. This dataset is primarily applied in fields including emotion research, behavioral psychology, security and communications, aiming to address the challenges in facial micro-expression and macro-expression recognition.

- 1SAMM Long Videos: A Spontaneous Facial Micro- and Macro-Expressions Dataset曼彻斯特都会大学计算机与数学系 · 2020年



