# Box2Seg: Learning Semantics of 3D Point Clouds with Box-Level Supervision

arXiv: https://arxiv.org/abs/2201.02963 · HTML: https://mlsanity.com/paper/2201.02963/

## Key facts

- Cited by 11 papers in the map (data through 14 Oct 2026).
- Builds on 47 of its 63 references that are in the map.
- First posted on arXiv 9 Jan 2022.

## Builds on (most cited first, 30 of 47)

- [Adam: A Method for Stochastic Optimization](https://mlsanity.com/paper/1412.6980/) (2014)
- [U-Net: Convolutional Networks for Biomedical Image Segmentation](https://mlsanity.com/paper/1505.04597/) (2015)
- [PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation](https://mlsanity.com/paper/1612.00593/) (2016)
- [nuScenes: A multimodal dataset for autonomous driving](https://mlsanity.com/paper/1903.11027/) (2019)
- [PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Space](https://mlsanity.com/paper/1706.02413/) (2017)
- [ScanNet: Richly-annotated 3D Reconstructions of Indoor Scenes](https://mlsanity.com/paper/1702.04405/) (2017)
- [ShapeNet: An Information-Rich 3D Model Repository](https://mlsanity.com/paper/1512.03012/) (2015)
- [Scalability in Perception for Autonomous Driving: Waymo Open Dataset](https://mlsanity.com/paper/1912.04838/) (2019)
- [Learning Deep Features for Discriminative Localization](https://mlsanity.com/paper/1512.04150/) (2015)
- [4D Spatio-Temporal ConvNets: Minkowski Convolutional Neural Networks](https://mlsanity.com/paper/1904.08755/) (2019)
- [KPConv: Flexible and Deformable Convolution for Point Clouds](https://mlsanity.com/paper/1904.08889/) (2019)
- [Temporal Ensembling for Semi-Supervised Learning](https://mlsanity.com/paper/1610.02242/) (2016)
- [PointCNN: Convolution On X-Transformed Points](https://mlsanity.com/paper/1801.07791/) (2018)
- [3D Semantic Segmentation with Submanifold Sparse Convolutional Networks](https://mlsanity.com/paper/1711.10275/) (2017)
- [PointConv: Deep Convolutional Networks on 3D Point Clouds](https://mlsanity.com/paper/1811.07246/) (2018)
- [RandLA-Net: Efficient Semantic Segmentation of Large-Scale Point Clouds](https://mlsanity.com/paper/1911.11236/) (2019)
- [Deep Learning for 3D Point Clouds: A Survey](https://mlsanity.com/paper/1912.12033/) (2019)
- [PartNet: A Large-scale Benchmark for Fine-grained and Hierarchical Part-level 3D Object Understanding](https://mlsanity.com/paper/1812.02713/) (2018)
- [PointContrast: Unsupervised Pre-training for 3D Point Cloud Understanding](https://mlsanity.com/paper/2007.10985/) (2020)
- [Searching Efficient 3D Architectures with Sparse Point-Voxel Convolution](https://mlsanity.com/paper/2007.16100/) (2020)
- [Joint 2D-3D-Semantic Data for Indoor Scene Understanding](https://mlsanity.com/paper/1702.01105/) (2017)
- [Cylindrical and Asymmetrical 3D Convolution Networks for LiDAR Segmentation](https://mlsanity.com/paper/2011.10033/) (2020)
- [Large-scale Point Cloud Semantic Segmentation with Superpoint Graphs](https://mlsanity.com/paper/1711.09869/) (2017)
- [Point-Voxel CNN for Efficient 3D Deep Learning](https://mlsanity.com/paper/1907.03739/) (2019)
- [SqueezeSeg: Convolutional Neural Nets with Recurrent CRF for Real-Time Road-Object Segmentation from 3D LiDAR Point Cloud](https://mlsanity.com/paper/1710.07368/) (2017)
- [SqueezeSegV2: Improved Model Structure and Unsupervised Domain Adaptation for Road-Object Segmentation from a LiDAR Point Cloud](https://mlsanity.com/paper/1809.08495/) (2018)
- [Exploring Data-Efficient 3D Scene Understanding with Contrastive Scene Contexts](https://mlsanity.com/paper/2012.09165/) (2020)
- [SEGCloud: Semantic Segmentation of 3D Point Clouds](https://mlsanity.com/paper/1710.07563/) (2017)
- [OccuSeg: Occupancy-aware 3D Instance Segmentation](https://mlsanity.com/paper/2003.06537/) (2020)
- [Tangent Convolutions for Dense Prediction in 3D](https://mlsanity.com/paper/1807.02443/) (2018)

## Built on by (most cited first, 11 of 11)

- [PartSLIP: Low-Shot Part Segmentation for 3D Point Clouds via Pretrained Image-Language Models](https://mlsanity.com/paper/2212.01558/) (2022)
- [LaserMix for Semi-Supervised LiDAR Semantic Segmentation](https://mlsanity.com/paper/2207.00026/) (2022)
- [Multi-Modal Data-Efficient 3D Scene Understanding for Autonomous Driving](https://mlsanity.com/paper/2405.05258/) (2024)
- [LiDAL: Inter-frame Uncertainty Based Active Learning for 3D LiDAR Semantic Segmentation](https://mlsanity.com/paper/2211.05997/) (2022)
- [Beyond One Shot, Beyond One Perspective: Cross-View and Long-Horizon Distillation for Better LiDAR Representations](https://mlsanity.com/paper/2507.05260/) (2025)
- [U4D: Uncertainty-Aware 4D World Modeling from LiDAR Sequences](https://mlsanity.com/paper/2512.02982/) (2025)
- [One Class One Click: Quasi Scene-level Weakly Supervised Point Cloud Semantic Segmentation with Active Learning](https://mlsanity.com/paper/2211.12657/) (2022)
- [ItTakesTwo: Leveraging Peer Representations for Semi-supervised LiDAR Semantic Segmentation](https://mlsanity.com/paper/2407.07171/) (2024)
- [LiDAR Remote Sensing Meets Weak Supervision: Concepts, Methods, and Perspectives](https://mlsanity.com/paper/2503.18384/) (2025)
- [BoxCell: Leveraging SAM for Cell Segmentation with Box Supervision](https://mlsanity.com/paper/2311.17960/) (2023)
- [PinPoint3D: Fine-Grained 3D Part Segmentation from a Few Clicks](https://mlsanity.com/paper/2509.25970/) (2025)

## Abstract (arXiv)

Learning dense point-wise semantics from unstructured 3D point clouds with fewer labels, although a realistic problem, has been under-explored in literature. While existing weakly supervised methods can effectively learn semantics with only a small fraction of point-level annotations, we find that the vanilla bounding box-level annotation is also informative for semantic segmentation of large-scale 3D point clouds. In this paper, we introduce a neural architecture, termed Box2Seg, to learn point-level semantics of 3D point clouds with bounding box-level supervision. The key to our approach is to generate accurate pseudo labels by exploring the geometric and topological structure inside and outside each bounding box. Specifically, an attention-based self-training (AST) technique and Point Class Activation Mapping (PCAM) are utilized to estimate pseudo-labels. The network is further trained and refined with pseudo labels. Experiments on two large-scale benchmarks including S3DIS and ScanNet demonstrate the competitive performance of the proposed method. In particular, the proposed network can be trained with cheap, or even off-the-shelf bounding box-level annotations and subcloud-level tags.