adirik / wonder3d

Generates 3D assets from images

  • Public
  • 2.3K runs
  • GitHub
  • Paper
  • License

Input

Output

Run time and cost

This model runs on Nvidia A40 (Large) GPU hardware.

Readme

Wonder3D

Wonder3D is an image-to-3D model that can generate 3D assets in a few minutes. See the original repository and paper for details. Note that this model is intended to generate singular objects and can not generate scenes.

How to use the API

To use Wonder3D, simply upload an (ideally) front-facing image of the object you’d like to create a 3D asset for. You can remove the image background manually before uploading or enable the remove_background option to automatically detect and remove the image background. The API input arguments are as follows:

  • image: image to generate a 3D asset from.
  • remove_background: whether to detect and remove image background, set to True by default. Only set it to False if you are uploading an image with removed background.
  • num_steps: number of optimization steps.
  • seed: seed for reproducibility, default value is None. Set to an arbitrary value for deterministic generation.

The API returns a textured mesh in .glb format.

References

@misc{long2023wonder3d,
      title={Wonder3D: Single Image to 3D using Cross-Domain Diffusion}, 
      author={Xiaoxiao Long and Yuan-Chen Guo and Cheng Lin and Yuan Liu and Zhiyang Dou and Lingjie Liu and Yuexin Ma and Song-Hai Zhang and Marc Habermann and Christian Theobalt and Wenping Wang},
      year={2023},
      eprint={2310.15008},
      archivePrefix={arXiv},
      primaryClass={cs.CV}
}