# Unsupervised Primitive Discovery for Improved 3D Generative Modeling

arXiv: https://arxiv.org/abs/1906.03650 · HTML: https://mlsanity.com/paper/1906.03650/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 9 Jun 2019.

## Built on by (most cited first, 8 of 8)

- [Neural Shape Mating: Self-Supervised Object Assembly with Adversarial Shape Priors](https://mlsanity.com/paper/2205.14886/) (2022)
- [Multi-modal Machine Learning in Engineering Design: A Review and Future Directions](https://mlsanity.com/paper/2302.10909/) (2023)
- [3D Geometric Shape Assembly via Efficient Point Cloud Matching](https://mlsanity.com/paper/2407.10542/) (2024)
- [Depth from Camera Motion and Object Detection](https://mlsanity.com/paper/2103.01468/) (2021)
- [Primitive3D: 3D Object Dataset Synthesis from Randomly Assembled Primitives](https://mlsanity.com/paper/2205.12627/) (2022)
- [TACO-Net: Topological Signatures Triumph in 3D Object Classification](https://mlsanity.com/paper/2509.24802/) (2025)
- [3D GANs and Latent Space: A comprehensive survey](https://mlsanity.com/paper/2304.03932/) (2023)
- [LiLa-Net: Lightweight Latent LiDAR Autoencoder for 3D Point Cloud Reconstruction](https://mlsanity.com/paper/2510.02028/) (2025)

## Abstract (arXiv)

3D shape generation is a challenging problem due to the high-dimensional output space and complex part configurations of real-world objects. As a result, existing algorithms experience difficulties in accurate generative modeling of 3D shapes. Here, we propose a novel factorized generative model for 3D shape generation that sequentially transitions from coarse to fine scale shape generation. To this end, we introduce an unsupervised primitive discovery algorithm based on a higher-order conditional random field model. Using the primitive parts for shapes as attributes, a parameterized 3D representation is modeled in the first stage. This representation is further refined in the next stage by adding fine scale details to shape. Our results demonstrate improved representation ability of the generative model and better quality samples of newly generated 3D shapes. Further, our primitive generation approach can accurately parse common objects into a simplified representation.