3RL

3RL数据集是由阿拉伯国际大学创建的面部表情识别数据集，包含约24,394张图像，这些图像来自三个著名数据集的合并与编辑。数据集标记了五种基本情绪：快乐、恐惧、悲伤、厌恶和愤怒。创建过程中，研究人员手动检查并删除了误导性图像，确保了数据集的质量。3RL数据集主要用于面部表情识别研究，旨在提高情绪分类的准确性和泛化能力。

The 3RL dataset is a facial expression recognition dataset created by Arab International University. It contains approximately 24,394 images compiled and edited from three renowned datasets. The dataset is annotated with five basic emotions: happiness, fear, sadness, disgust and anger. During its development, researchers manually inspected and removed misleading images to ensure the dataset's quality. The 3RL dataset is primarily used for facial expression recognition research, aiming to improve the accuracy and generalization ability of emotion classification.