One-2-3-45: Any Single Image to 3D Mesh in 45 Seconds without Per-Shape Optimization

1UC San Diego, 2UCLA, 3Cornell University, 4Zhejiang University, 5IIT Madras, 6Adobe Research
* Equal contribution

Abstract

Single image 3D reconstruction is an important but challenging task that requires extensive knowledge of our natural world. Many existing methods solve this problem by optimizing a neural radiance field under the guidance of 2D diffusion models but suffer from lengthy optimization time, 3D inconsistency results, and poor geometry. In this work, we propose a novel method that takes a single image of any object as input and generates a full 360-degree 3D textured mesh in a single feed-forward pass. Given a single image, we first use a view-conditioned 2D diffusion model, Zero123, to generate multi-view images for the input view, and then aim to lift them up to 3D space. Since traditional reconstruction methods struggle with inconsistent multi-view predictions, we build our 3D reconstruction module upon an SDF-based generalizable neural surface reconstruction method and propose several critical training strategies to enable the reconstruction of 360-degree meshes. Without costly optimizations, our method reconstructs 3D shapes in significantly less time than existing methods. Moreover, our method favors better geometry, generates more 3D consistent results, and adheres more closely to the input image. We evaluate our approach on both synthetic data and in-the-wild images and demonstrate its superiority in terms of both mesh quality and runtime. In addition, our approach can seamlessly support the text-to-3D task by integrating with off-the-shelf text-to-image diffusion models.

Method

Our method consists of three primary components: (a) Multi-view synthesis: we use a view-conditioned 2D diffusion model, Zero123 [36], to generate multi-view images in a two-stage manner. The input of Zero123 includes a single image and a relative camera transformation, which is parameterized by the relative spherical coordinates (∆θ, ∆φ, ∆r). (b) Pose estimation: we estimate the elevation angle θ of the input image based on four nearby views generated by Zero123. We then obtain the poses of the multi-view images by combining the specified relative poses with the estimated pose of the input view. (c) 3D reconstruction: We feed the multi-view posed images to an SDF-based generalizable neural surface reconstruction module for 360◦ mesh reconstruction.

Text to 3D

We utilize DALL-E 2 to generate an image conditioned on the text and then lift it to 3D.

Comparison with Existing Methods




Image-to-3D Comparison.


Text-to-3D Comparison.


More comparison with Shap-E on real-world images. Row 2-3: One-2-3-45; Row 4-5: Shap-E.

BibTeX

@article{liu2023one,
  title={One-2-3-45: Any single image to 3d mesh in 45 seconds without per-shape optimization},
  author={Liu, Minghua and Xu, Chao and Jin, Haian and Chen, Linghao and Xu, Zexiang and Su, Hao and others},
  journal={arXiv preprint arXiv:2306.16928},
  year={2023}
}