论文标题
无监督的运动运动检测，用于部分分段的3D形状集合
Unsupervised Kinematic Motion Detection for Part-segmented 3D Shape Collections
论文作者
论文摘要
制造物体的3D模型对于填充虚拟世界和视觉和机器人技术的合成数据很重要。为了最有用，应该阐明此类对象：它们的部分应在与之互动时移动。尽管存在铰接式对象数据集，但创建它们是劳动密集型的。基于学习的零件动作预测可以有所帮助，但是所有现有方法都需要带注释的培训数据。在本文中，我们提出了一种无监督的方法，用于发现部分分段的3D形状集合中的铰接运动。我们的方法是基于一个概念，我们称之为封闭类别：对物体部分的任何有效表达都应将对象保留在同一语义类别中（例如，椅子保持椅子）。我们使用一种算法来实现此概念，该算法优化了形状的零件运动参数，从而可以转换为集合中的其他形状。我们通过使用Partnet-Mobility数据集重新发现零件动作来评估我们的方法。对于几乎所有形状类别，我们方法的预测运动参数在地面真实注释方面的错误较低，表现优于两种监督运动预测方法。
3D models of manufactured objects are important for populating virtual worlds and for synthetic data generation for vision and robotics. To be most useful, such objects should be articulated: their parts should move when interacted with. While articulated object datasets exist, creating them is labor-intensive. Learning-based prediction of part motions can help, but all existing methods require annotated training data. In this paper, we present an unsupervised approach for discovering articulated motions in a part-segmented 3D shape collection. Our approach is based on a concept we call category closure: any valid articulation of an object's parts should keep the object in the same semantic category (e.g. a chair stays a chair). We operationalize this concept with an algorithm that optimizes a shape's part motion parameters such that it can transform into other shapes in the collection. We evaluate our approach by using it to re-discover part motions from the PartNet-Mobility dataset. For almost all shape categories, our method's predicted motion parameters have low error with respect to ground truth annotations, outperforming two supervised motion prediction methods.
