跨视角基于学习的关系对象匹配（英文版）

英文标题：Learning-based Relational Object Matching Across Views中文摘要：本论文提出了一种基于 RGB 图像的目标检测匹配的学习方法，结合了局部关键点与新颖的物体级特征，通过关联图神经网络训练，在许多视图上具有较好的性能并优于现有的纯关键点匹配方法。英文摘要：Intelligent robots require object-level scene understanding to reason aboutpossible tasks and interactions with the environment. Moreover, many perceptiontasks such as scene reconstruction, image retrieval, or place recognition canbenefit from reasoning on the level of objects. While keypoint-based matchingcan yield strong results for findin