What question did this study set out to answer?

This work aims to improve 3D shape generation using probabilistic diffusion models and neural signed distance functions.

October 1, 2023

Diffusion-SDF: Conditional Generative Modeling of Signed Distance Functions

View Full Paper

Ask AI

Bookmark

Authors

Randomized trial demonstrates effective shape completion and reconstruction in 3D, indicating new possibilities for generative models.

This work aims to improve 3D shape generation using probabilistic diffusion models and neural signed distance functions.
Proposed Diffusion-SDF model for shape completion and single-view reconstruction.
Utilized neural signed distance functions to represent 3D geometry from various inputs.
Developed a custom modulation module for learning the reversal of neural network weights.
Demonstrated realistic unconditional generation of 3D shapes.
Achieved effective conditional generation from partial inputs in extensive experiments.

Chou et al. (2023) studied this question.

View Full Paper

Ask AI

Bookmark

October 1, 2023

View Full Paper

Ask AI

Bookmark

Authors

Randomized trial demonstrates effective shape completion and reconstruction in 3D, indicating new possibilities for generative models.

This work aims to improve 3D shape generation using probabilistic diffusion models and neural signed distance functions.
Proposed Diffusion-SDF model for shape completion and single-view reconstruction.
Utilized neural signed distance functions to represent 3D geometry from various inputs.
Developed a custom modulation module for learning the reversal of neural network weights.
Demonstrated realistic unconditional generation of 3D shapes.
Achieved effective conditional generation from partial inputs in extensive experiments.

Chou et al. (2023) studied this question.

View Full Paper

Ask AI

Bookmark