EUVIMEDEUVIMED entdecken
Crossref

Motion-Guided Disentanglement for Point Cloud Masked Autoencoders

Haoran Wang, Shaqing Song, Baosheng Yu, Tong Jia, Dongyue Chen, Chunfeng Yuan, Weiming Hu, Haibin Ling

IEEE Transactions on Circuits and Systems for Video Technology · 2026-08 · 36 (8) · 10887-10897

DOI: 10.1109/tcsvt.2026.3675668 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Transactions on Circuits and Systems for Video Technology, 2026-08. Verfasst von Haoran Wang und 7 weiteren Autor:innen. Die Arbeit nennt 60 Literaturangaben. Gefördert von: Fundamental Research Funds for the Central Universities of China, Open Projects Program of State Key Laboratory of Multimodal Artificial Intelligence Systems, Research Fund of Liaoning Province.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Masked autoencoders have been extended beyond images, but random masking often fails to capture non-uniform motion regions in inherently disordered and irregular data like point cloud videos. In this paper, we propose a motion-guided disentanglement method (MGD) to improve masked autoencoders for point cloud video representation learning. Specifically, we begin by estimating motion intensity using an optimal transport approach, which guides the separate masking of dynamic and static regions. This motion-guided masking ensures balanced coverage, addressing the limitations of random masking in capturing non-uniformly distributed motion regions. Furthermore, we disentangle the prediction tasks into motion prediction for high-motion point tubes and appearance reconstruction for low-motion ones. This disentanglement enables the model to more effectively capture both motion and appearance in point cloud videos. We conducted experiments on four widely used point cloud video datasets—NTU RGB+D, MSR-Action3D, NvGesture, and SHREC’17—which demonstrate that our approach consistently improves masked autoencoders for point cloud video representation learning, achieving new state-of-the-art results. Code will be publicly available on GitHub.

Förderung

  • Fundamental Research Funds for the Central Universities of China
  • Open Projects Program of State Key Laboratory of Multimodal Artificial Intelligence Systems
  • Research Fund of Liaoning Province

Zitieren

APA

Wang, H., Song, S., Yu, B., Jia, T., Chen, D., Yuan, C., Hu, W., & Ling, H. (2026). Motion-Guided Disentanglement for Point Cloud Masked Autoencoders. IEEE Transactions on Circuits and Systems for Video Technology, 36(8), 10887-10897. https://doi.org/10.1109/tcsvt.2026.3675668

Vancouver

Wang H, Song S, Yu B, Jia T, Chen D, Yuan C, et al. Motion-Guided Disentanglement for Point Cloud Masked Autoencoders. IEEE Transactions on Circuits and Systems for Video Technology. 2026;36(8):10887-10897. doi:10.1109/tcsvt.2026.3675668

BibTeX
@article{Wang2026,
  author = {Wang, Haoran and Song, Shaqing and Yu, Baosheng and Jia, Tong and Chen, Dongyue and Yuan, Chunfeng and Hu, Weiming and Ling, Haibin},
  title = {Motion-Guided Disentanglement for Point Cloud Masked Autoencoders},
  journal = {IEEE Transactions on Circuits and Systems for Video Technology},
  year = {2026},
  volume = {36},
  number = {8},
  pages = {10887--10897},
  doi = {10.1109/tcsvt.2026.3675668}
}
Kostenlos starten →