BoxSplitGen: A Generative Model for 3D Part Bounding Boxes in Varying Granularity

TL;DR

This paper introduces BoxSplitGen, a novel generative framework for creating 3D shapes by iteratively splitting bounding boxes to refine shape details, aiding human-like 3D design processes.

Contribution

We propose a new generative model that produces 3D part bounding boxes with varying granularity and integrates with 3D diffusion models for shape generation.

Findings

BoxSplitGen outperforms token prediction and inpainting models.

The box-to-shape model achieves superior results with state-of-the-art diffusion.

Framework enables intuitive 3D shape refinement from coarse to fine.

Abstract

Human creativity follows a perceptual process, moving from abstract ideas to finer details during creation. While 3D generative models have advanced dramatically, models specifically designed to assist human imagination in 3D creation -- particularly for detailing abstractions from coarse to fine -- have not been explored. We propose a framework that enables intuitive and interactive 3D shape generation by iteratively splitting bounding boxes to refine the set of bounding boxes. The main technical components of our framework are two generative models: the box-splitting generative model and the box-to-shape generative model. The first model, named BoxSplitGen, generates a collection of 3D part bounding boxes with varying granularity by iteratively splitting coarse bounding boxes. It utilizes part bounding boxes created through agglomerative merging and learns the reverse of the merging process -- the splitting sequences. The model consists of two main components: the first learns the categorical distribution of the box to be split, and the second learns the distribution of the two new boxes, given the set of boxes and the indication of which box to split. The second model, the box-to-shape generative model, is trained by leveraging the 3D shape priors learned by an existing 3D diffusion model while adapting the model to incorporate bounding box conditioning. In our experiments, we demonstrate that the box-splitting generative model outperforms token prediction models and the inpainting approach with an unconditional diffusion model. Also, we show that our box-to-shape model, based on a state-of-the-art 3D diffusion model, provides superior results compared to a previous model.