论文标题
OSOP：多阶段的一击对象姿势估计框架
OSOP: A Multi-Stage One Shot Object Pose Estimation Framework
论文作者
论文摘要
我们提出了一种用于对象检测的新颖的单发方法和6个DOF姿势估计，该方法不需要对目标对象进行培训。在测试时，它将目标图像和纹理3D查询模型作为输入。核心想法是代表一个3D模型，该模型具有从不同观点呈现的许多2D模板。这使得基于CNN的直接密集特征提取和匹配。该对象首先定位在2D中，然后估计其近似观点，然后进行致密的2d-3d对应性预测。最终姿势是用PNP计算的。我们评估了素摩德，遮挡，自制，YCB-V和Tless数据集的方法，并报告与对合成数据培训的最新方法相比，即使我们的方法未接受用于测试的对象模型的培训，也与对合成数据的最新方法相比。
We present a novel one-shot method for object detection and 6 DoF pose estimation, that does not require training on target objects. At test time, it takes as input a target image and a textured 3D query model. The core idea is to represent a 3D model with a number of 2D templates rendered from different viewpoints. This enables CNN-based direct dense feature extraction and matching. The object is first localized in 2D, then its approximate viewpoint is estimated, followed by dense 2D-3D correspondence prediction. The final pose is computed with PnP. We evaluate the method on LineMOD, Occlusion, Homebrewed, YCB-V and TLESS datasets and report very competitive performance in comparison to the state-of-the-art methods trained on synthetic data, even though our method is not trained on the object models used for testing.
