Recognizing 3D part instances from a 3D point cloud is crucial for 3D structure and scene understanding. Several learning-based approaches use semantic segmentation and instance center prediction as training tasks and fail to further exploit the inherent relationship between shape semantics and part instances. In this paper, we present a new method for 3D part instance segmentation. Our method exploits semantic segmentation to fuse nonlocal instance features, such as center prediction, and further enhances the fusion scheme in a multi- and cross-level way. We also propose a semantic region center prediction task to train and leverage the prediction results to improve the clustering of instance points. Our method outperforms existing methods with a large-margin improvement in the PartNet benchmark. We also demonstrate that our feature fusion scheme can be applied to other existing methods to improve their performance in indoor scene instance segmentation tasks.

Oil-impregnated porous polyimide (iPPI) materials are usually used as retainer for bearings. In these bearings, balls and rings, balls and retainers are two different kinds of contact. In this paper, the friction and wear properties of iPPI were investigated using steel (disc)—steel (ball)—iPPI (pin) double-contact friction test rig for simulating the actual contact in bearings. The results show that compared with that of iPPI—steel single contact, the friction coefficient of iPPI—steel in double contacts is lower and decreases with the amount of additional oil. The surface of iPPI in single contact suffers more wear compared with that in double contacts. Different from single contact, the worn surfaces of iPPI in double contacts are blackened. The Raman spectra of worn surfaces of balls and discs indicate that α-Fe2O3 and Fe3O4 were formed during rubbing of the double contacts. Many nanoscale iron oxide particles are found on the worn surfaces of iPPI in double contacts; on the contrary, few particles could be found on the surface in single contact. In double-contact friction, the nanoscale wear debris penetrates inside the iPPI material through the process of extruding and recycling of oil, which is the mechanism of the blackening of the iPPI worn surfaces. The studies show that the double-contact friction method is a new and effective method to study the friction in bearings, especially for those with polymer retainer.