Unsupervised Primitive Discovery for Improved 3D Generative Modeling
This addresses the problem of accurate generative modeling of 3D shapes for computer vision and graphics applications, representing an incremental improvement over existing methods.
The paper tackles the challenge of 3D shape generation by proposing a factorized generative model that transitions from coarse to fine scale generation, using an unsupervised primitive discovery algorithm based on a higher-order conditional random field model. The results demonstrate improved representation ability and better quality samples of newly generated 3D shapes.
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.