# H-CNN: Spatial Hashing Based CNN for 3D Shape Analysis

arXiv: https://arxiv.org/abs/1803.11385 · HTML: https://mlsanity.com/paper/1803.11385/

## Key facts

- Cited by 10 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 30 Mar 2018.

## Built on by (most cited first, 10 of 10)

- [OctFormer: Octree-based Transformers for 3D Point Clouds](https://mlsanity.com/paper/2305.03045/) (2023)
- [Dual Octree Graph Networks for Learning Adaptive Volumetric Shape Representations](https://mlsanity.com/paper/2205.02825/) (2022)
- [A Closer Look at Few-Shot 3D Point Cloud Classification](https://mlsanity.com/paper/2303.18210/) (2023)
- [Zero-Shot Multi-Object Scene Completion](https://mlsanity.com/paper/2403.14628/) (2024)
- [NFANet: A Novel Method for Weakly Supervised Water Extraction from High-Resolution Remote Sensing Imagery](https://mlsanity.com/paper/2201.03686/) (2022)
- [VSFormer: Mining Correlations in Flexible View Set for Multi-view 3D Shape Understanding](https://mlsanity.com/paper/2409.09254/) (2024)
- [Interpolation-Aware Padding for 3D Sparse Convolutional Neural Networks](https://mlsanity.com/paper/2108.06925/) (2021)
- [Two Heads are Better than One: Geometric-Latent Attention for Point Cloud Classification and Segmentation](https://mlsanity.com/paper/2111.00231/) (2021)
- [Dynamic Local Feature Aggregation for Learning on Point Clouds](https://mlsanity.com/paper/2301.02836/) (2023)
- [2L-LSH: A Locality-Sensitive Hash Function-Based Method For Rapid Point Cloud Indexing](https://mlsanity.com/paper/2604.21442/) (2026)

## Abstract (arXiv)

We present a novel spatial hashing based data structure to facilitate 3D shape analysis using convolutional neural networks (CNNs). Our method well utilizes the sparse occupancy of 3D shape boundary and builds hierarchical hash tables for an input model under different resolutions. Based on this data structure, we design two efficient GPU algorithms namely hash2col and col2hash so that the CNN operations like convolution and pooling can be efficiently parallelized. The spatial hashing is nearly minimal, and our data structure is almost of the same size as the raw input. Compared with state-of-the-art octree-based methods, our data structure significantly reduces the memory footprint during the CNN training. As the input geometry features are more compactly packed, CNN operations also run faster with our data structure. The experiment shows that, under the same network structure, our method yields comparable or better benchmarks compared to the state-of-the-art while it has only one-third memory consumption. Such superior memory performance allows the CNN to handle high-resolution shape analysis.