3D shape of people from monocular video with SelfRecon
3D shape of people from monocular video with SelfRecon
SelfRecon: Self Reconstruction Your Digital Avatar from Monocular Video
arXiv paper abstract https://arxiv.org/abs/2201.12792
arXiv PDF paper https://arxiv.org/pdf/2201.12792.pdf
Project page https://jby1993.github.io/SelfRecon
We propose SelfRecon, a clothed human body reconstruction method that combines implicit and explicit representations to recover space-time coherent geometries from a monocular self-rotating human video.
Explicit methods require a predefined template mesh … while the template is hard to acquire … fixed topology limits … accuracy and clothing types.
Implicit methods support arbitrary topology and have high quality … However, … difficult to integrate multi-frame information to produce a consistent registration
… propose to combine the advantages of both representations.
… utilize differential mask loss of the explicit mesh to obtain the coherent overall shape, while the details on the implicit surface are refined with the differentiable neural rendering. … explicit mesh is updated … to adjust its topology changes
… SelfRecon can produce high-fidelity surfaces for arbitrary clothed humans … on real captured monocular videos.
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
