ShapeShifter: 3D Variations Using Multiscale and Sparse Point-Voxel Diffusion

1Inria, 2Adobe Research, 3École polytechnique
MY ALT TEXT

Abstract

This paper proposes a new 3D generative model that learns to synthesize shape variations based on a single example. While generative methods for 3D objects have recently attracted much attention, current techniques often lack geometric details and/or require long training times and large resources. Our approach remedies these issues by combining sparse voxel grids and multiscale point, normal, and color sampling within an encoder-free neural architecture that can be trained efficiently and in parallel. We show that our resulting variations better capture the fine details of their original input and can capture more general types of surfaces than previous SDF-based methods. Moreover, we offer interactive generation of 3D shape variants, allowing more human control in the design loop if needed.

Results

Input Geometry

Sin3DM

ShapeShifter (ours)

ShapeShifter generates high-quality 3D shape variations based on a single input shape. Its unique representation allows for efficient and detailed 3D shape generation.

Multiscale Diffusion

MY ALT TEXT

3D shape representation. We operate on raw features: oriented point clouds (possibly augmented with RGB colors) encoded on sparse voxel grids for efficient computations.

Multiscale diffusion on sparse voxel grid. We start from noise at the coarsest level, and obtain the 3D feature grid through reverse diffusion. Each subsequent level uses the output of the previous level. Inactive voxels are first pruned, then upsampled with a level-specific upsampler. The upsampled grid is subsequently noised and passed through the diffusion model to obtain a clean version of the sparse feature grid. All levels are independent and can thus be trained in parallel.

Acknowledgements

This work was supported by 3IA Côte d'Azur (ANR-19-P3IA-0002).