Share your thoughts, 1 month free Claude Pro on usSee more
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
Semantic Scene CompletionSemanticKITTI (test)--
67
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
Point cloud generationShapeNet Car
1-NNA (CD)54.55
41
3D Shape GenerationShapeNet airplane
1-NNA (CD)73.82
30
Point Cloud CompletionShapeNet (test)
EMD (Airplane)1.03
26
Point cloud generationShapeNet chair
1-NNA (CD)57.09
23
3D Shape GenerationShapeNet chair
1-NN Accuracy (CD)56.26
18
Point cloud generationShapeNet Chair (test)
1-NNA (CD)54.6
16
Showing 10 of 43 rows

Other info

Follow for update