Get 3D object shape from sparse views using epipolar features to guide diffusion model with Sparse3D
Get 3D object shape from sparse views using epipolar features to guide diffusion model with Sparse3D
Sparse3D: Distilling Multiview-Consistent Diffusion for Object Reconstruction from Sparse Views
arXiv paper abstract https://arxiv.org/abs/2308.14078
arXiv PDF paper https://arxiv.org/pdf/2308.14078.pdf
Reconstructing 3D objects from extremely sparse views is a long-standing and challenging problem ... image diffusion ... struggle to simultaneously achieve high-quality ... for ... novel-view synthesis (NVS) and geometry.
... present Sparse3D, a novel 3D reconstruction method tailored for sparse view inputs ... distills robust priors from a multiview-consistent diffusion model to refine a neural radiance field.
... employ a controller that harnesses epipolar features from input views, guiding a pre-trained diffusion model, such as Stable Diffusion, to produce novel-view images that maintain 3D consistency with the input.
By tapping into 2D priors from powerful image diffusion models, ... integrated model consistently delivers high-quality results, even when faced with open-world objects.
To address the blurriness introduced by conventional SDS, ... introduce the category-score distillation sampling (C-SDS) to enhance detail.
... approach outperforms previous state-of-the-art works on the metrics regarding NVS and geometry reconstruction.
Please like and share this post if you enjoyed it using the buttons at the bottom!
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
コメント