3DGS-Avatar: Animatable Avatars via Deformable 3D Gaussian Splatting

arXiv:2312.09228v3 Announce Type: replace
Abstract: We introduce an approach that creates animatable human avatars from monocular videos using 3D Gaussian Splatting (3DGS). Existing methods based on neural radiance fields (NeRFs) achieve high-quality novel-view/novel-pose image synthesis but often require days of training, and are extremely slow at inference time. Recently, the community has explored fast grid structures for efficient training of clothed avatars. Albeit being extremely fast at training, these methods can barely achieve an interactive rendering frame rate with around …

abstract arxiv avatar avatars community cs.cv fields human image inference neural radiance fields novel quality synthesis training type via videos view