Learning Object Bounding Boxes for 3D Instance Segmentation on Point Clouds

We propose a novel, conceptually simple and general framework for instance segmentation on 3D point clouds. Our method, called 3D-BoNet, follows the simple design philosophy of per-point multilayer perceptrons (MLPs). The framework directly regresses 3D bounding boxes for all instances in a point cloud, while simultaneously predicting a point-level mask for each instance. It consists of a backbone network followed by two parallel network branches for 1) bounding box regression and 2) point mask prediction. 3D-BoNet is single-stage, anchor-free and end-to-end trainable. Moreover, it is remarkably computationally efficient as, unlike existing approaches, it does not require any post-processing steps such as non-maximum suppression, feature sampling, clustering or voting. Extensive experiments show that our approach surpasses existing work on both ScanNet and S3DIS datasets while being approximately 10x more computationally efficient. Comprehensive ablation studies demonstrate the effectiveness of our design.

PDF Abstract NeurIPS 2019 PDF NeurIPS 2019 Abstract

Results from the Paper


Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
3D Instance Segmentation S3DIS 3D-BoNet mRec 47.6 # 13
mPrec 65.6 # 12
3D Instance Segmentation ScanNet(v2) 3D-BoNet mRec 47.6 # 2
mAP 25.3 # 14
mAP @ 50 48.8 # 15

Methods


No methods listed for this paper. Add relevant methods here