HoloFusion: Towards Photo-realistic 3D Generative Modeling

TL;DR

HoloFusion is a novel 3D generative modeling approach that combines diffusion techniques and super-resolution to produce high-fidelity, view-consistent 3D models from multi-view 2D images, surpassing existing methods in realism.

Contribution

The paper introduces HoloFusion, a new method that integrates coarse 3D generation, multi-view super-resolution, and distillation into a unified framework for realistic 3D modeling.

Findings

Achieves the most realistic results on CO3Dv2 dataset.

Outperforms baselines like DreamFusion and HoloDiffusion.

Produces high-fidelity, view-consistent 3D samples.

Abstract

Diffusion-based image generators can now produce high-quality and diverse samples, but their success has yet to fully translate to 3D generation: existing diffusion methods can either generate low-resolution but 3D consistent outputs, or detailed 2D views of 3D objects but with potential structural defects and lacking view consistency or realism. We present HoloFusion, a method that combines the best of these approaches to produce high-fidelity, plausible, and diverse 3D samples while learning from a collection of multi-view 2D images only. The method first generates coarse 3D samples using a variant of the recently proposed HoloDiffusion generator. Then, it independently renders and upsamples a large number of views of the coarse 3D model, super-resolves them to add detail, and distills those into a single, high-fidelity implicit 3D representation, which also ensures view consistency of the final renders. The super-resolution network is trained as an integral part of HoloFusion, end-to-end, and the final distillation uses a new sampling scheme to capture the space of super-resolved signals. We compare our method against existing baselines, including DreamFusion, Get3D, EG3D, and HoloDiffusion, and achieve, to the best of our knowledge, the most realistic results on the challenging CO3Dv2 dataset.