CrowdSim2
收藏资源简介:
CrowdSim2是由西里西亚工业大学开发的用于人和车辆检测的合成数据集,基于Unity图形引擎的模拟器生成。该数据集包含数千个视频片段,涵盖多种天气条件和对象密度,用于精确的对象定位。数据集的创建过程自动化,无需人工标注,有效降低了成本。CrowdSim2主要应用于计算机视觉领域,特别是在需要复杂场景和隐私保护的数据标注任务中,为深度学习模型提供了一个受控的测试环境。
CrowdSim2 is a synthetic dataset for human and vehicle detection, developed by Silesian University of Technology and generated via a simulator based on the Unity graphics engine. This dataset comprises thousands of video clips spanning diverse weather conditions and object densities, tailored for accurate object localization. The entire dataset creation process is fully automated, eliminating the need for manual annotation and effectively reducing costs. CrowdSim2 is primarily utilized in the field of computer vision, particularly in data annotation tasks requiring complex scenarios and privacy protection, serving as a controlled test environment for deep learning models.




