Get 3D object shape using a prior from fusing multiple local SDF fields with PR-NeuS
Get 3D object shape using a prior from fusing multiple local SDF fields with PR-NeuS
PR-NeuS: A Prior-based Residual Learning Paradigm for Fast Multi-view Neural Surface Reconstruction
arXiv paper abstract https://arxiv.org/abs/2312.11577
arXiv PDF paper https://arxiv.org/pdf/2312.11577.pdf
Neural surfaces learning has ... impressive performance in multi-view surface reconstruction. However, most ... use large multilayer perceptrons (MLPs) resulting in hours of training
... propose a prior-based residual learning paradigm for ... reconstruction. This paradigm consists of two optimization stages.
... first ... propose to leverage generalization models to generate a basis signed distance function (SDF) field. This ... can ... be ... obtained by fusing multiple local SDF fields produced by generalization models. This provides a coarse global geometry prior.
... second ... a fast residual learning strategy based on hash-encoding networks is proposed to encode an offset SDF field for the basis SDF field.
... introduce a prior-guided sampling scheme to help the residual learning stage converge better, and thus recover finer structures.
... takes about 3 minutes to reconstruct the surface of a single scene, while achieving competitive surface quality ...
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
Comentarios