Abstract
In recent years, 3D point cloud semantic segmentation has made remarkable progress. However, most existing work focuses on designing intricate structures to aggregate local features, resulting in a significant number of parameters and computational demands. In this Paper, we combine the idea of structural reparameterization in 2D convolution to propose RPNet, which can effectively reduce the number of parameters while fully extracting the point cloud features. Specifically. We first design the multi-branch structure SRLFA (Structure Re-parameterization Local Feature Abstract) module based on the reparameterization to fully extract the local features of the point cloud, and design the PFA (Point Feature Abstract) module to extract the features of the point itself. Then, by decoupling the training and inference phases, the multi-branch structure is fused into an equivalent single-branch structure through the idea of structural reparameterization during training and inference, which ensures the feature extraction capability while effectively reducing the number of parameters. Finally, the proposed method is trained and tested on several public data, and the results demonstrate that the proposed method achieves advanced performance in mIoU and OA with effective control of the number of model parameters.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Qi, C.R., Su, H., Mo, K., Guibas, L.J.: PointNet: deep learning on point sets for 3D classification and segmentation (2017). http://arxiv.org/abs/1612.00593
Qi, C.R., Yi, L., Su, H., Guibas, L.J.: PointNet++: deep hierarchical feature learning on point sets in a metric space, p. 14 (2017)
Thomas, H., Qi, C.R., Deschaud, J.-E., Marcotegui, B., Goulette, F., Guibas, L.: KPConv: flexible and deformable convolution for point clouds. In: 2019 IEEE/CVF International Conference on Computer Vision (ICCV), pp. 6410–6419. IEEE, Seoul, Korea (South) (2019). https://doi.org/10.1109/ICCV.2019.00651
Xu, M., Ding, R., Zhao, H., Qi, X.: PAConv: position adaptive convolution with dynamic kernel assembling on point clouds. In: 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp. 3172–3181. IEEE, Nashville, TN, USA (2021). https://doi.org/10.1109/CVPR46437.2021.00319
Graham, B., Engelcke, M., van der Maaten, L.: 3D semantic segmentation with submanifold sparse convolutional networks. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (2018)
Qiu, S., Anwar, S., Barnes, N.: Semantic segmentation for real point cloud scenes via bilateral augmentation and adaptive fusion (2021). http://arxiv.org/abs/2103.07074
Singh, G., Gupta, S., Lease, M., Dawson, C.: Range-net: a high precision streaming SVD for big data applications. arXiv preprint arXiv:2010.14226 (2020)
Wei, X., Yu, R., Sun, J.: View-GCN: view-based graph convolutional network for 3D shape analysis. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (2020)
Riegler, G., Osman Ulusoy, A., Geiger, A.: OctNet: learning deep 3D representations at high resolutions. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (2017)
Phan, A.V., Nguyen, M.L., Nguyen, Y.L.H., Bui, L.T.: DGCNN: a convolutional neural network over large-scale labeled graphs. Neural Netw.Netw. 108, 533–543 (2018). https://doi.org/10.1016/j.neunet.2018.09.001
Li, G., Muller, M., Thabet, A., Ghanem, B.: DeepGCNs: can GCNs go as deep as CNNs? In: Proceedings of the IEEE/CVF International Conference on Computer Vision (2019)
Ding, X., Zhang, X., Ma, N., Han, J., Ding, G., Sun, J.: RepVGG: making VGG-style ConvNets great again (2021). http://arxiv.org/abs/2101.03697
Ding, X., Zhang, X., Han, J., Ding, G.: Diverse branch block: building a convolution as an inception-like unit (2021). http://arxiv.org/abs/2103.13425, https://doi.org/10.48550/arXiv.2103.13425
Armeni, I., Sax, A., Zamir, A.R., Savarese, S.: Joint 2D-3D-semantic data for indoor scene understanding (2017)
Dai, A., Chang, A.X., Savva, M., Halber, M., Funkhouser, T., Niessner, M.: ScanNet: richly-annotated 3D reconstructions of indoor scenes. In: 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 2432–2443. IEEE, Honolulu, HI (2017). https://doi.org/10.1109/CVPR.2017.261
Chang, A.X., et al.: ShapeNet: an information-rich 3D model repository (2015). http://arxiv.org/abs/1512.03012, https://doi.org/10.48550/arXiv.1512.03012
Li, Y., Bu, R., Sun, M., Wu, W., Di, X., Chen, B.: PointCNN: convolution on x-transformed points. In: Advances in Neural Information Processing Systems. Curran Associates, Inc. (2018)
Hu, Q., et al.: RandLA-net: efficient semantic segmentation of large-scale point clouds (2020). http://arxiv.org/abs/1911.11236
Zhao, H., Jiang, L., Jia, J., Torr, P.H.S., Koltun, V.: Point Transformer. https://doi.org/10.1109/ICCV48922.2021.01595
Tang, L., Zhan, Y., Chen, Z., Yu, B., Tao, D.: Contrastive boundary learning for point cloud segmentation. In: 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp. 8479–8489. IEEE, New Orleans, LA, USA (2022). https://doi.org/10.1109/CVPR52688.2022.00830
Qian, G., et al.: PointNeXt: revisiting PointNet++ with improved training and scaling strategies (2022). http://arxiv.org/abs/2206.04670
Lin, H., et al.: Meta architecture for point cloud analysis (2022). http://arxiv.org/abs/2211.14462
Xiang, T., Zhang, C., Song, Y., Yu, J., Cai, W.: Walk in the cloud: learning curves for point clouds shape analysis (2021). http://arxiv.org/abs/2105.01288
Qian, G., Hammoud, H.A.A.K., Li, G., Thabet, A., Ghanem, B.: ASSANet: an anisotropic separable set abstraction for efficient point cloud representation learning (2021). http://arxiv.org/abs/2110.10538
Lai, X., et al.: Stratified transformer for 3D point cloud segmentation. https://doi.org/10.1109/CVPR52688.2022.00831
Acknowledgement
This research was funded by the Basic Research Program of Qinghai Province (Grant No. 2021-ZJ-704) and Beijing Natural Science Foundation (GrantNo. 4212001).
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2023 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Li, Z., Jia, K., Zhao, Y., Huang, W. (2023). Structural Reparameterization Network on Point Cloud Semantic Segmentation. In: Lu, H., et al. Image and Graphics. ICIG 2023. Lecture Notes in Computer Science, vol 14355. Springer, Cham. https://doi.org/10.1007/978-3-031-46305-1_28
Download citation
DOI: https://doi.org/10.1007/978-3-031-46305-1_28
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-46304-4
Online ISBN: 978-3-031-46305-1
eBook Packages: Computer ScienceComputer Science (R0)