# Adaptive O-CNN: A Patch-based Deep Representation of 3D Shapes

arXiv: https://arxiv.org/abs/1809.07917 · HTML: https://mlsanity.com/paper/1809.07917/

## Key facts

- Cited by 25 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 21 Sep 2018.

## Built on by (most cited first, 25 of 25)

- [Neural Kernel Surface Reconstruction](https://mlsanity.com/paper/2305.19590/) (2023)
- [FlatFormer: Flattened Window Attention for Efficient Point Cloud Transformer](https://mlsanity.com/paper/2301.08739/) (2023)
- [SA-ConvONet: Sign-Agnostic Optimization of Convolutional Occupancy Networks](https://mlsanity.com/paper/2105.03582/) (2021)
- [3D Shape Reconstruction from 2D Images with Disentangled Attribute Flow](https://mlsanity.com/paper/2203.15190/) (2022)
- [Towards High-Fidelity Single-view Holistic Reconstruction of Indoor Scenes](https://mlsanity.com/paper/2207.08656/) (2022)
- [Octree Transformer: Autoregressive 3D Shape Generation on Hierarchically Structured Sequences](https://mlsanity.com/paper/2111.12480/) (2021)
- [Neural Template: Topology-aware Reconstruction and Disentangled Generation of 3D Meshes](https://mlsanity.com/paper/2206.04942/) (2022)
- [Enable Deep Learning on Mobile Devices: Methods, Systems, and Applications](https://mlsanity.com/paper/2204.11786/) (2022)
- [OcTr: Octree-based Transformer for 3D Object Detection](https://mlsanity.com/paper/2303.12621/) (2023)
- [OctGPT: Octree-based Multiscale Autoregressive Models for 3D Shape Generation](https://mlsanity.com/paper/2504.09975/) (2025)
- [ME-PCN: Point Completion Conditioned on Mask Emptiness](https://mlsanity.com/paper/2108.08187/) (2021)
- [PVNAS: 3D Neural Architecture Search with Point-Voxel Convolution](https://mlsanity.com/paper/2204.11797/) (2022)
- [Learning Mesh Representations via Binary Space Partitioning Tree Networks](https://mlsanity.com/paper/2106.14274/) (2021)
- [Learning Body-Aware 3D Shape Generative Models](https://mlsanity.com/paper/2112.07022/) (2021)
- [A QuadTree Image Representation for Computational Pathology](https://mlsanity.com/paper/2108.10873/) (2021)
- [A Review of Deep Learning-Powered Mesh Reconstruction Methods](https://mlsanity.com/paper/2303.02879/) (2023)
- [Deep Mesh Prior: Unsupervised Mesh Restoration using Graph Convolutional Networks](https://mlsanity.com/paper/2107.02909/) (2021)
- [ActivationNet: Representation learning to predict contact quality of interacting 3-D surfaces in engineering designs](https://mlsanity.com/paper/2103.11288/) (2021)
- [Interpolation-Aware Padding for 3D Sparse Convolutional Neural Networks](https://mlsanity.com/paper/2108.06925/) (2021)
- [Traffic Event Detection as a Slot Filling Problem](https://mlsanity.com/paper/2109.06035/) (2021)
- [Intelligent Generation of Graphical Game Assets: A Conceptual Framework and Systematic Review of the State of the Art](https://mlsanity.com/paper/2311.10129/) (2023)
- [DiffPoint: Single and Multi-view Point Cloud Reconstruction with ViT Based Diffusion Model](https://mlsanity.com/paper/2402.11241/) (2024)
- [NASM: Neural Anisotropic Surface Meshing](https://mlsanity.com/paper/2410.23109/) (2024)
- [Reconstructing vehicles from orthographic drawings using deep neural networks](https://mlsanity.com/paper/2206.08789/) (2022)
- [INRet: A General Framework for Accurate Retrieval of INRs for Shapes](https://mlsanity.com/paper/2501.15722/) (2025)

## Abstract (arXiv)

We present an Adaptive Octree-based Convolutional Neural Network (Adaptive O-CNN) for efficient 3D shape encoding and decoding. Different from volumetric-based or octree-based CNN methods that represent a 3D shape with voxels in the same resolution, our method represents a 3D shape adaptively with octants at different levels and models the 3D shape within each octant with a planar patch. Based on this adaptive patch-based representation, we propose an Adaptive O-CNN encoder and decoder for encoding and decoding 3D shapes. The Adaptive O-CNN encoder takes the planar patch normal and displacement as input and performs 3D convolutions only at the octants at each level, while the Adaptive O-CNN decoder infers the shape occupancy and subdivision status of octants at each level and estimates the best plane normal and displacement for each leaf octant. As a general framework for 3D shape analysis and generation, the Adaptive O-CNN not only reduces the memory and computational cost, but also offers better shape generation capability than the existing 3D-CNN approaches. We validate Adaptive O-CNN in terms of efficiency and effectiveness on different shape analysis and generation tasks, including shape classification, 3D autoencoding, shape prediction from a single image, and shape completion for noisy and incomplete point clouds.