SurFhead: Affine Rig Blending for Geometrically Accurate 2D Gaussian
Surfel-based Head Avatars

Jaeseong Lee*1, Taewoong Kang*1, Marcel Bühler2, Minjung Kim1, Sungwon Hwang1, Junha Hyung1, Hyojin Jang1, Jaegul Choo1
1KAIST, 2ETH Zürich
Paper OpenReview Code
Responsive image
SurFhead not only reconstructs photorealistic head avatars but also high-fidelity surface normals, depth, and meshes from RGB videos alone. These avatars are represented through affine rigging of 2D surfel splats bound to a parametric morphable face model. SurFhead allow us to fully control poses, expressions, and viewpoints, enhancing the both appearance and geometry.

Cross-reenactment results of SurFhead.

Left: held-out sequence; Right: Cross-reenactment.

Abstract

Recent advancements in head avatar rendering using Gaussian primitives have achieved significantly high-fidelity results. Although precise head geometry is crucial for applications like mesh reconstruction and relighting, current meth- ods struggle to capture intricate geometric details and render unseen poses due to their reliance on similarity transformations, which cannot handle stretch and shear transforms essential for detailed deformations of geometry. To address this, we propose SurFhead, a novel method that reconstructs riggable head geometry from RGB videos using 2D Gaussian surfels, which offer well-defined geometric properties, such as precise depth from fixed ray intersections and normals derived from their surface orientation, making them advantageous over 3D counterparts. SurFhead ensures high-fidelity rendering of both normals and images, even in ex- treme poses, by leveraging classical mesh-based deformation transfer and affine transformation interpolation. SurFhead introduces precise geometric deforma- tion and blends surfels through polar decomposition of transformations, including those affecting normals. Our key contribution lies in bridging classical graphics techniques, such as mesh-based deformation, with modern Gaussian primitives, achieving state-of-the-art geometry reconstruction and rendering quality. Unlike previous avatar rendering approaches, SurFhead enables efficient reconstruction driven by Gaussian primitives while preserving high-fidelity geometry.

Method overview

Responsive image
Overall pipeline of SurFhead. Only from RGB videos, SurFhead constructs geometrically accurate head avatars, equipped with our intricate deformations. The Jacobian covers stretch and shear deformations avoiding surface distortion. Moreover, the blended Jacobian alleviates inherent local deformations’ discontinuity. Finally, elaborated modeling of eyeballs such as preser- vation of specularity and convexity achieves more realistic appearance and geometry.

Self-reenactment comparison with baselines.

Cross-reenactment comparison with baselines.

Results for relighting with GaussianShader.

BibTeX

If you find this work useful, please consider citing:

@inproceedings{
	lee2025surfhead,
	title={SurFhead: Affine Rig Blending for Geometrically Accurate 2D Gaussian Surfel Head Avatars},
	author={Jaeseong Lee and Taewoong Kang and Marcel Buehler and Min-Jung Kim and Sungwon Hwang and Junha Hyung and Hyojin Jang and Jaegul Choo},
	booktitle={The Thirteenth International Conference on Learning Representations},
	year={2025},
	url={https://openreview.net/forum?id=1x1gGg49jr}
	}
}