SynPick
收藏资源简介:
SynPick是由德国波恩大学自主智能系统组创建的一个合成数据集,专注于动态场景理解,特别是在工业应用中的箱子拣选场景。该数据集包含503,232条记录,涵盖了动态拣选序列,旨在解决6D姿态估计和跟踪问题。数据集的创建过程涉及使用NVIDIA PhysX物理引擎进行对象排列生成和操作模拟。SynPick数据集的应用领域包括机器人抓取、状态估计和预测,特别是在工业和服务机器人领域。
SynPick is a synthetic dataset developed by the Autonomous Intelligent Systems Group at the University of Bonn, Germany, focusing on dynamic scene understanding, especially bin picking scenarios in industrial applications. This dataset consists of 503,232 records covering dynamic picking sequences, and is intended to solve 6D pose estimation and tracking problems. The dataset creation process utilizes the NVIDIA PhysX physics engine for object arrangement generation and manipulation simulation. The application fields of the SynPick dataset include robotic grasping, state estimation and prediction, particularly in industrial and service robotics scenarios.




