SVIRO
收藏资源简介:
SVIRO数据集由德国人工智能研究中心创建,专注于车辆内部后座乘客的检测与分类。该数据集包含10种不同车辆内部场景,共计25000条数据,涵盖了对象检测、实例分割、姿态估计和深度图像等多种数据类型。数据集的创建过程中使用了Blender等3D图形软件,模拟真实环境中的光照和材质。SVIRO数据集的应用领域主要集中在自动驾驶和车辆安全系统,旨在通过机器学习模型提高对后座乘客状态的识别准确性和可靠性。
The SVIRO dataset was created by the German Research Centre for Artificial Intelligence, focusing on the detection and classification of rear-seat passengers inside vehicles. It includes 10 distinct vehicle interior scenarios, with a total of 25,000 data entries, covering multiple data modalities such as object detection, instance segmentation, pose estimation, and depth images. During the dataset's construction, 3D graphics software including Blender was utilized to simulate real-world lighting and materials. The SVIRO dataset is primarily applied in autonomous driving and vehicle safety systems, with the goal of enhancing the accuracy and reliability of rear-seat passenger state recognition through machine learning models.




