# ASAP-Net: Attention and Structure Aware Point Cloud Sequence Segmentation

arXiv: https://arxiv.org/abs/2008.05149 · HTML: https://mlsanity.com/paper/2008.05149/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 12 Aug 2020.

## Built on by (most cited first, 8 of 8)

- [Geometric Clifford Algebra Networks](https://mlsanity.com/paper/2302.06594/) (2023)
- [Spatial-Temporal Transformer for 3D Point Cloud Sequences](https://mlsanity.com/paper/2110.09783/) (2021)
- [SVQNet: Sparse Voxel-Adjacent Query Network for 4D Spatio-Temporal LiDAR Semantic Segmentation](https://mlsanity.com/paper/2308.13323/) (2023)
- [TASeg: Temporal Aggregation Network for LiDAR Semantic Segmentation](https://mlsanity.com/paper/2407.09751/) (2024)
- [Multi-body SE(3) Equivariance for Unsupervised Rigid Segmentation and Motion Estimation](https://mlsanity.com/paper/2306.05584/) (2023)
- [Semantics-Guided Moving Object Segmentation with 3D LiDAR](https://mlsanity.com/paper/2205.03186/) (2022)
- [SiamMo: Siamese Motion-Centric 3D Object Tracking](https://mlsanity.com/paper/2408.01688/) (2024)
- [Point Cloud Segmentation Using Sparse Temporal Local Attention](https://mlsanity.com/paper/2112.00289/) (2021)

## Abstract (arXiv)

Recent works of point clouds show that mulit-frame spatio-temporal modeling outperforms single-frame versions by utilizing cross-frame information. In this paper, we further improve spatio-temporal point cloud feature learning with a flexible module called ASAP considering both attention and structure information across frames, which we find as two important factors for successful segmentation in dynamic point clouds. Firstly, our ASAP module contains a novel attentive temporal embedding layer to fuse the relatively informative local features across frames in a recurrent fashion. Secondly, an efficient spatio-temporal correlation method is proposed to exploit more local structure for embedding, meanwhile enforcing temporal consistency and reducing computation complexity. Finally, we show the generalization ability of the proposed ASAP module with different backbone networks for point cloud sequence segmentation. Our ASAP-Net (backbone plus ASAP module) outperforms baselines and previous methods on both Synthia and SemanticKITTI datasets (+3.4 to +15.2 mIoU points with different backbones). Code is availabe at https://github.com/intrepidChw/ASAP-Net