EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC-ND 4.0

PointMamba++: Rethinking Ordering and Convolution Strategy of State Space Model for Point Cloud Analysis

Ke Xu, Xinpu Liu, Yinghui Gao, Qingyong Hu, Xinjie Wang, Hanyun Wang, Yulan Guo

IET Computer Vision · 2026-01 · 20 (1)

DOI: 10.1049/cvi2.70064 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IET Computer Vision, online erschienen am 06.04.2026. Verfasst von Ke Xu und 6 weiteren Autor:innen. Die Arbeit nennt 66 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0). Gefördert von: National Natural Science Foundation of China.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

With the linear complexity and long‐sequence global modelling capability, Mamba becomes a competitor to Transformer architectures in point clouds analysis. However, the designs of traditional 1D convolutions and reordering strategies, do not match the inherently unordered nature of point clouds, which constrain the performance enhancement. In this work, we rethink the ordering and convolution strategy of the PointMamba, and present a novel architecture named PointMamba++ to more effectively aggregate local structural features and achieve a superior accuracy and computation trade‐off. Specifically, we design a point‐edge convolution to aggregate neighbourhood features of point cloud tokens, which replaces 1D convolution layers in traditional Mamba modules and does not perform convolution by sequence but according to geometric relationships. Furthermore, considering that forcibly ordering point clouds is not conducive to learning local geometric features and easily leads to unstable sequence dependencies, we design a sequence‐independent BiMamba module, which adopts two reverse and parallel scanning paths, to reduce the dependency on sequential scanning of Mamba while enhancing point cloud representation abilities. Extensive experiments show that PointMamba++ surpasses typical convolution‐based and Transformer‐based architectures, and achieves state‐of‐the‐art performance on multiple tasks including shape classification, part segmentation, and semantic segmentation.

Förderung

  • National Natural Science Foundation of China

Zitieren

APA

Xu, K., Liu, X., Gao, Y., Hu, Q., Wang, X., Wang, H., & Guo, Y. (2026). PointMamba++: Rethinking Ordering and Convolution Strategy of State Space Model for Point Cloud Analysis. IET Computer Vision, 20(1). https://doi.org/10.1049/cvi2.70064

Vancouver

Xu K, Liu X, Gao Y, Hu Q, Wang X, Wang H, et al. PointMamba++: Rethinking Ordering and Convolution Strategy of State Space Model for Point Cloud Analysis. IET Computer Vision. 2026;20(1). doi:10.1049/cvi2.70064

BibTeX
@article{Xu2026,
  author = {Xu, Ke and Liu, Xinpu and Gao, Yinghui and Hu, Qingyong and Wang, Xinjie and Wang, Hanyun and Guo, Yulan},
  title = {PointMamba++: Rethinking Ordering and Convolution Strategy of State Space Model for Point Cloud Analysis},
  journal = {IET Computer Vision},
  year = {2026},
  volume = {20},
  number = {1},
  doi = {10.1049/cvi2.70064}
}
Kostenlos starten →