Implicit neural fields, typically encoded by a multilayer perceptron (MLP) that maps from coordinates (e.g., xyz) to signals (e.g., signed distances), have shown remarkable promise as a high-fidelity and compact representation. However, the lack of a regular and explicit grid structure also makes it challenging to apply generative modeling directly on implicit neural fields in order to synthesize new data.
We propose HyperDiffusion, a novel approach for unconditional generative modeling of implicit neural fields. HyperDiffusion operates directly on MLP weights and generates new neural implicit fields encoded by synthesized MLP parameters. Specifically, a collection of MLPs is first optimized to faithfully represent individual data samples. Subsequently, a diffusion process is trained in this MLP weight space to model the underlying distribution of neural implicit fields.
HyperDiffusion enables diffusion modeling over a implicit, compact, and yet high-fidelity representation of complex signals across 3D shapes and 4D mesh animations within one single unified framework.
@misc{erkoç2023hyperdiffusion,
title={HyperDiffusion: Generating Implicit Neural Fields with Weight-Space Diffusion},
author={Ziya Erkoç and Fangchang Ma and Qi Shan and Matthias Nießner and Angela Dai},
year={2023},
eprint={2303.17015},
archivePrefix={arXiv},
primaryClass={cs.CV}
}