Any3DIS: Class-Agnostic 3D Instance Segmentation by 2D Mask Tracking

Figure 1. Comparison of our proposed approach, Any3DIS, with existing 3D instance segmentation methods such as Open3DIS [19]. Open3DIS frequently encounters over-segmentation issues, generating redundant 3D proposals due to its unsupervised merging process. In contrast, our approach leverages robust guidance from 2D mask tracking to maintain consistent object segmentation across video frames, effectively enhancing segmentation accuracy and being 10 times faster than Open3DIS.