An Introduction to 3-D Sparse CNNs
IEEE Computer Graphics and Applications · 2026-07 · 46 (4) · 104-110
DOI: 10.1109/mcg.2026.3696340 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Computer Graphics and Applications, 2026-07. Verfasst von Peng-Shuai Wang und 2 weiteren Autor:innen. Die Arbeit nennt 7 Literaturangaben. Gefördert von: National Natural Science Foundation of China.
Abstract (Originalsprache)
Three-dimensional data are now central to computer vision, robotics, autonomous driving, medical imaging, augmented reality, and computer graphics. Yet, 3-D deep learning is difficult because geometric data are often large, irregular, and mostly empty. Dense voxel convolutional neural networks (CNNs) provide a simple extension of image CNNs to 3-D, but their memory and computation grow cubically with resolution. Sparse 3-D CNNs address this problem by storing and processing only the active parts of space, such as surface voxels or occupied cells in a point cloud. This article gives an accessible introduction to voxel-based CNNs, sparse convolution, and the data structures that make sparse 3-D networks practical, with special attention to octrees, hash tables, and the practical lessons and pitfalls that matter in real systems.
Förderung
- National Natural Science Foundation of China
Zitieren
APA
Wang, P. S., Pietroni, N., & Hu, R. (2026). An Introduction to 3-D Sparse CNNs. IEEE Computer Graphics and Applications, 46(4), 104-110. https://doi.org/10.1109/mcg.2026.3696340
Vancouver
Wang PS, Pietroni N, Hu R. An Introduction to 3-D Sparse CNNs. IEEE Computer Graphics and Applications. 2026;46(4):104-110. doi:10.1109/mcg.2026.3696340
BibTeX
@article{Wang2026,
author = {Wang, Peng-Shuai and Pietroni, Nico and Hu, Ruizhen},
title = {An Introduction to 3-D Sparse CNNs},
journal = {IEEE Computer Graphics and Applications},
year = {2026},
volume = {46},
number = {4},
pages = {104--110},
doi = {10.1109/mcg.2026.3696340}
}
EUVIMED