Get 3D shape of novel object from one image using pre-trained diffusion models with Zero-1-to-3
Get 3D shape of novel object from one image using pre-trained diffusion models with Zero-1-to-3
Zero-1-to-3: Zero-shot One Image to 3D Object
arXiv paper abstract https://arxiv.org/abs/2303.11328
arXiv PDF paper https://arxiv.org/pdf/2303.11328.pdf
… introduce Zero-1-to-3, a framework for changing the camera viewpoint of an object given just a single RGB image.
To perform novel view synthesis in this under-constrained setting, … capitalize on the geometric priors that large-scale diffusion models learn about natural images.
… conditional diffusion model uses a synthetic dataset to learn controls of the relative camera viewpoint, which allow new images to be generated of the same object under a specified camera transformation.
Even though it is trained on a synthetic dataset, … model retains a strong zero-shot generalization ability to out-of-distribution datasets as well as in-the-wild images
… viewpoint-conditioned diffusion approach can further be used for the task of 3D reconstruction from a single image.
… method significantly outperforms state-of-the-art single-view 3D reconstruction and novel view synthesis models by leveraging Internet-scale pre-training.
Stay up to date. Subscribe to my posts https://morrislee1234.wixsite.com/website/contact
Web site with my other posts by category https://morrislee1234.wixsite.com/website
LinkedIn https://www.linkedin.com/in/morris-lee-47877b7b
