XAILab-CyberSpark/Cabin-Human-Behavior-Dataset

「CyberData塞塔」智能座舱用户行为数据集是一个专为加速智能座舱感知算法开发而设计的高质量、程序化生成的图像数据集。它包含5000张高保真座舱内用户行为图像，涵盖不同年龄、性别、种族和衣着风格的虚拟人模型，以及多种驾驶与乘坐行为和面部表情。数据集可直接用于智能座舱端侧视觉模型，尤其适用于DMS/OMS算法的训练、微调与验证。

The "CyberData Saita" In-Cabin User Behavior Dataset is a high-quality, programmatically generated image dataset designed to accelerate the development of in-cabin perception algorithms. It includes 5,000 photo-realistic images of in-cabin user behaviors, covering virtual human models of various ages, genders, ethnicities, and clothing styles, as well as a variety of driving and riding behaviors and facial expressions. The dataset is directly applicable to on-device vision models in smart cockpits, especially for training, fine-tuning, and validation of DMS/OMS algorithms.