Scalable 3D Panoptic Segmentation As Superpoint Graph Clustering

Abstract

We introduce a highly efficient method for panoptic segmentation of large 3D point clouds by redefining this task as a scalable graph clustering problem. This approach can be trained using only local auxiliary tasks, thereby eliminating the resource-intensive instance-matching step during training. Moreover, our formulation can easily be adapted to the superpoint paradigm, further increasing its efficiency. This allows our model to process scenes with millions of points and thousands of objects in a single inference. Our method, called SuperCluster, achieves a new state-of-the-art panoptic segmentation performance for two indoor scanning datasets: $50.1$ PQ ($+7.8$) for S3DIS Area~5, and $58.7$ PQ ($+25.2$) for ScanNetV2. We also set the first state-of-the-art for two large-scale mobile mapping benchmarks: KITTI-360 and DALES. With only $209$k parameters, our model is over $30$ times smaller than the best-competing method and trains up to $15$ times faster. Our code and pretrained models are available at https://github.com/drprojects/superpoint_transformer.

arXiv commentAccepted at 3DV 2024, Oral presentation

References in corpus (8)

• PointNeXt: Revisiting PointNet++ with Improved Training and Scaling Strategies
• Learning Object Bounding Boxes for 3D Instance Segmentation on Point Clouds
• Point Transformer V2: Grouped Vector Attention and Partition-based Pooling
• OctFormer: Octree-based Transformers for 3D Point Clouds
• Deep Learning Based 3D Segmentation: A Survey
• KITTI-360: A Novel Dataset and Benchmarks for Urban Scene Understanding in 2D and 3D
• Window Normalization: Enhancing Point Cloud Understanding by Unifying Inconsistent Point Densities
• Parallel Cut Pursuit For Minimization of the Graph Total Variation