Reproduce “MGAvatar: Mesh-Bound Gaussians for Head Avatar Geometry and Appearance Modeling” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Lei Shi, Sen Peng, Zhiyang Deng, Zhonggui Chen, Xiaohu Guo, Baorong Yang, Xiao Dong
MGAvatar reconstructs animatable heads from video by first using constrained, vertex-bound Gaussians to refine a FLAME mesh, then switching to face-bound Gaussians for appearance. Identity offsets and pose-dependent deformation let the mesh include regions such as hair and shoulders while retaining FLAME topology. On NeRSemble novel-view synthesis, it reports 34.570 PSNR versus STAvatar’s 33.803, but does not resolve every structure.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.