GeoAvatar: Adaptive Geometrical Gaussian Splatting for 3D Head Avatar

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Moon, SeungJun, Lew, Hah Min, Lee, Seungeun, Kang, Ji-Su, Park, Gyeong-Moon
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866913958841024512
author Moon, SeungJun
Lew, Hah Min
Lee, Seungeun
Kang, Ji-Su
Park, Gyeong-Moon
author_facet Moon, SeungJun
Lew, Hah Min
Lee, Seungeun
Kang, Ji-Su
Park, Gyeong-Moon
contents Despite recent progress in 3D head avatar generation, balancing identity preservation, i.e., reconstruction, with novel poses and expressions, i.e., animation, remains a challenge. Existing methods struggle to adapt Gaussians to varying geometrical deviations across facial regions, resulting in suboptimal quality. To address this, we propose GeoAvatar, a framework for adaptive geometrical Gaussian Splatting. GeoAvatar leverages Adaptive Pre-allocation Stage (APS), an unsupervised method that segments Gaussians into rigid and flexible sets for adaptive offset regularization. Then, based on mouth anatomy and dynamics, we introduce a novel mouth structure and the part-wise deformation strategy to enhance the animation fidelity of the mouth. Finally, we propose a regularization loss for precise rigging between Gaussians and 3DMM faces. Moreover, we release DynamicFace, a video dataset with highly expressive facial motions. Extensive experiments show the superiority of GeoAvatar compared to state-of-the-art methods in reconstruction and novel animation scenarios.
format Preprint
id arxiv_https___arxiv_org_abs_2507_18155
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle GeoAvatar: Adaptive Geometrical Gaussian Splatting for 3D Head Avatar
Moon, SeungJun
Lew, Hah Min
Lee, Seungeun
Kang, Ji-Su
Park, Gyeong-Moon
Graphics
Computer Vision and Pattern Recognition
Machine Learning
Despite recent progress in 3D head avatar generation, balancing identity preservation, i.e., reconstruction, with novel poses and expressions, i.e., animation, remains a challenge. Existing methods struggle to adapt Gaussians to varying geometrical deviations across facial regions, resulting in suboptimal quality. To address this, we propose GeoAvatar, a framework for adaptive geometrical Gaussian Splatting. GeoAvatar leverages Adaptive Pre-allocation Stage (APS), an unsupervised method that segments Gaussians into rigid and flexible sets for adaptive offset regularization. Then, based on mouth anatomy and dynamics, we introduce a novel mouth structure and the part-wise deformation strategy to enhance the animation fidelity of the mouth. Finally, we propose a regularization loss for precise rigging between Gaussians and 3DMM faces. Moreover, we release DynamicFace, a video dataset with highly expressive facial motions. Extensive experiments show the superiority of GeoAvatar compared to state-of-the-art methods in reconstruction and novel animation scenarios.
title GeoAvatar: Adaptive Geometrical Gaussian Splatting for 3D Head Avatar
topic Graphics
Computer Vision and Pattern Recognition
Machine Learning
url https://arxiv.org/abs/2507.18155