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.

翻译：我们提出了一种高效的大规模三维点云全景分割方法，将这一任务重新定义为可扩展的图聚类问题。该方法仅需通过局部辅助任务进行训练，从而消除了训练过程中资源密集型的实例匹配步骤。此外，我们的方案可轻松适配超点范式，进一步提升了效率。这使得模型能够在单次推理中处理包含数百万个点和数千个对象的场景。我们提出的方法SuperCluster在两个室内扫描数据集上取得了最新的全景分割性能：在S3DIS Area~5上达到$50.1$ PQ（$+7.8$），在ScanNetV2上达到$58.7$ PQ（$+25.2$）。我们还首次在两个大规模移动测绘基准测试（KITTI-360和DALES）上建立了最优结果。该模型仅包含$209$k个参数，比最佳竞品模型小$30$倍以上，训练速度提升高达$15$倍。我们的代码与预训练模型已在https://github.com/drprojects/superpoint_transformer 开源。