Semantic Segmentation-Assisted Instance Feature Fusion for Multi-Level 3D Part Instance Segmentation
This addresses 3D structure understanding for robotics or AR/VR applications, but is incremental as it builds on existing segmentation and center prediction approaches.
The paper tackles 3D part instance segmentation by proposing a method that fuses semantic segmentation with instance features like center prediction in a multi-level way, and introduces a semantic region center prediction task to improve clustering. It achieves large-margin improvement on the PartNet benchmark and shows the fusion scheme can boost other methods in indoor scene tasks.
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.