Feed-forward Gaussian Registration for Head Avatar Creation and Editing
Malte Prinzler, Paulo F. U. Gotardo, Siyu Tang, Timo Bolkart
Abstract
We present MATCH (Multi-view Avatars from Topologically Corresponding Heads), a multi-view Gaussian registration method for high-quality head avatar creation and editing. State-of-the-art multi-view head avatar methods require time-consuming head tracking followed by expensive avatar optimization, often resulting in a total creation time of more than one day. MATCH, in contrast, directly predicts Gaussian splat textures in correspondence from calibrated multi-view images in just 0.5 seconds per frame, without requiring data preprocessing. The learned intra-subject correspondence across frames enables fast creation of personalized head avatars, while correspondence across subjects supports applications such as expression transfer, optimization-free tracking, semantic editing, and identity interpolation. We establish these correspondences end-to-end using a transformer-based model that predicts Gaussian splat textures in the fixed UV layout of a template mesh. To achieve this, we introduce a novel registration-guided attention block, where each UV-map token attends exclusively to image tokens depicting its corresponding mesh region. This design improves efficiency and performance compared to dense cross-view attention. MATCH outperforms existing methods in novel-view synthesis, geometry registration, and head avatar generation, while making avatar creation 10 times faster than the closest competing baseline. The code and model weights are available on the project website.
Ask Lune
Know this paper in minutes, not an afternoon.
Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.
Lineage
Builds on32
- 3D Gaussian Splatting for Real-Time Radiance Field RenderingBernhard Kerbl, Georgios Kopanas, Thomas Leimkühler, George DrettakisSIGGRAPH 2023 · 5,687 citations
- Learning an animatable detailed 3D face model from in-the-wild imagesYao Feng, Haiwen Feng, Michael J. Black, Timo BolkartSIGGRAPH 2021 · 662 citations
- 2D Gaussian Splatting for Geometrically Accurate Radiance FieldsBinbin Huang, Zehao Yu, Anpei Chen, Andreas Geiger et al.SIGGRAPH 2024 · 660 citations
- DUSt3R: Geometric 3D Vision Made EasyShuzhe Wang, Vincent Leroy, Yohann Cabon, Boris Chidlovskii et al.CVPR 2024 · 302 citations
- Mixture of volumetric primitives for efficient neural renderingStephen Lombardi, Tomas Simon, Gabriel Schwartz, Michael Zollhöfer et al.SIGGRAPH 2021 · 240 citations
Related papers
- PERSE: Personalized 3D Generative Avatars from A Single PortraitHyunsoo Cha, Inhee Lee, Hanbyul JooCVPR 2025
- FastAvatar: Towards Unified and Fast 3D Avatar Reconstruction with Large Gaussian Reconstruction TransformersYue Wu, Xuanhong Chen, Yufan Wu, Wen Li et al.ICLR 2026 · 7 citations
- FastGHA: Generalized Few-Shot 3D Gaussian Head Avatars with Real-Time AnimationXinya Ji, Sebastian Weiss, Manuel Kansy, Jacek Naruniec et al.ICLR 2026 · 6 citations
- SVG-Head: Hybrid Surface-Volumetric Gaussians for High-Fidelity Head Reconstruction and Real-Time EditingHeyi Sun, Cong Wang, Tian-Xing Xu, Jingwei Huang et al.ICCV 2025 · 8 citations
- GaussianAvatars: Photorealistic Head Avatars with Rigged 3D GaussiansShenhan Qian, Tobias Kirschstein, Liam Schoneveld, Davide Davoli et al.CVPR 2024 · 175 citations
