Magic123: One Image to High-Quality 3D Object Generation Using Both 2D and 3D Diffusion Priors
Guocheng Qian, Jinjie Mai, Abdullah Hamdi, Jian Ren, Aliaksandr Siarohin, Bing Li, Hsin-Ying Lee, Ivan Skorokhodov
Abstract
We present ``Magic123'', a two-stage coarse-to-fine approach for high-quality, textured 3D mesh generation from a single image in the wild using *both 2D and 3D priors*. In the first stage, we optimize a neural radiance field to produce a coarse geometry. In the second stage, we adopt a memory-efficient differentiable mesh representation to yield a high-resolution mesh with a visually appealing texture. In both stages, the 3D content is learned through reference-view supervision and novel-view guidance by a joint 2D and 3D diffusion prior. We introduce a trade-off parameter between the 2D and 3D priors to control the details and 3D consistencies of the generation. Magic123 demonstrates a significant improvement over previous image-to-3D techniques, as validated through extensive experiments on diverse synthetic and real-world images.
BibTeX
@inproceedings{
qian2024magic,
title={Magic123: One Image to High-Quality 3D Object Generation Using Both 2D and 3D Diffusion Priors},
author={Guocheng Qian and Jinjie Mai and Abdullah Hamdi and Jian Ren and Aliaksandr Siarohin and Bing Li and Hsin-Ying Lee and Ivan Skorokhodov and Peter Wonka and Sergey Tulyakov and Bernard Ghanem},
booktitle={The Twelfth International Conference on Learning Representations},
year={2024},
url={https://openreview.net/forum?id=0jHkUDyEO9}
}