Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

3D Shape Generation and Completion through Point-Voxel Diffusion

About

We propose a novel approach for probabilistic generative modeling of 3D shapes. Unlike most existing models that learn to deterministically translate a latent vector to a shape, our model, Point-Voxel Diffusion (PVD), is a unified, probabilistic formulation for unconditional shape generation and conditional, multi-modal shape completion. PVD marries denoising diffusion models with the hybrid, point-voxel representation of 3D shapes. It can be viewed as a series of denoising steps, reversing the diffusion process from observed point cloud data to Gaussian noise, and is trained by optimizing a variational lower bound to the (conditional) likelihood function. Experiments demonstrate that PVD is capable of synthesizing high-fidelity shapes, completing partial point clouds, and generating multiple completion results from single-view depth scans of real objects.

Linqi Zhou, Yilun Du, Jiajun Wu• 2021

Related benchmarks

TaskDatasetResultRank
3D point cloud generationShapeNet Car (test)
1-NNA (CD)64.49
57
3D point cloud generationShapeNet Chair category (test)
MMD (CD)2.622
56
3D point cloud generationShapeNet Airplane category (test)
1-NNA (CD, %)73.82
55
Semantic Scene CompletionSemanticKITTI (test)
Overall IoU21.2
48
Point cloud generationShapeNet Car
1-NNA (CD)54.55
27
Point cloud generationShapeNet chair
1-NNA (CD)57.09
23
3D Shape GenerationShapeNet airplane
1-NNA (CD)73.82
16
Point cloud generationShapeNet Chair (test)
1-NNA (CD)54.6
16
3D Scene CompletionSemanticKITTI (val)
JSD (BEV)0.498
14
Point cloud generationShapeNetPart Airplane (test)
1-NNA (CD)73.8
13
Showing 10 of 36 rows

Other info

Follow for update