_version_ 1866918488913739776
author Ntavelis, Evangelos
Wu, Sean
Shahbazi, Mohamad
Maninchedda, Fabio
Kostiaev, Dmitry
Sevastopolsky, Artem
Megaro, Vittorio
Phillips, Trevor
Blumentals, Alejandro
Ravikumar, Shridhar
Gupta, Mehak
Knothe, Reinhard
Bayer, Jeronimo
Vestner, Matthias
Schaefer, Simon
Etterlin, Thomas
Zimmermann, Christian
Deschler, Mathias
Kaufmann, Peter
Brugger, Stefan
Martin, Sebastian
Amberg, Brian
Runia, Tom
author_facet Ntavelis, Evangelos
Wu, Sean
Shahbazi, Mohamad
Maninchedda, Fabio
Kostiaev, Dmitry
Sevastopolsky, Artem
Megaro, Vittorio
Phillips, Trevor
Blumentals, Alejandro
Ravikumar, Shridhar
Gupta, Mehak
Knothe, Reinhard
Bayer, Jeronimo
Vestner, Matthias
Schaefer, Simon
Etterlin, Thomas
Zimmermann, Christian
Deschler, Mathias
Kaufmann, Peter
Brugger, Stefan
Martin, Sebastian
Amberg, Brian
Runia, Tom
contents We propose HeadsUp, a scalable feed-forward method for reconstructing high-quality 3D Gaussian heads from large-scale multi-camera setups. Our method employs an efficient encoder-decoder architecture that compresses input views into a compact latent representation. This latent representation is then decoded into a set of UV-parameterized 3D Gaussians anchored to a neutral head template. This UV representation decouples the number of 3D Gaussians from the number and resolution of input images, enabling training with many high-resolution input views. We train and evaluate our model on an internal dataset with more than 10,000 subjects, which is an order of magnitude larger than existing multi-view human head datasets. HeadsUp achieves state-of-the-art reconstruction quality and generalizes to novel identities without test-time optimization. We extensively analyze the scaling behavior of our model across identities, views, and model capacity, revealing practical insights for quality-compute trade-offs. Finally, we highlight the strength of our latent space by showcasing two downstream applications: generating novel 3D identities and animating the 3D heads with expression blendshapes.
format Preprint
id arxiv_https___arxiv_org_abs_2605_04035
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Large-Scale High-Quality 3D Gaussian Head Reconstruction from Multi-View Captures
Ntavelis, Evangelos
Wu, Sean
Shahbazi, Mohamad
Maninchedda, Fabio
Kostiaev, Dmitry
Sevastopolsky, Artem
Megaro, Vittorio
Phillips, Trevor
Blumentals, Alejandro
Ravikumar, Shridhar
Gupta, Mehak
Knothe, Reinhard
Bayer, Jeronimo
Vestner, Matthias
Schaefer, Simon
Etterlin, Thomas
Zimmermann, Christian
Deschler, Mathias
Kaufmann, Peter
Brugger, Stefan
Martin, Sebastian
Amberg, Brian
Runia, Tom
Computer Vision and Pattern Recognition
Machine Learning
We propose HeadsUp, a scalable feed-forward method for reconstructing high-quality 3D Gaussian heads from large-scale multi-camera setups. Our method employs an efficient encoder-decoder architecture that compresses input views into a compact latent representation. This latent representation is then decoded into a set of UV-parameterized 3D Gaussians anchored to a neutral head template. This UV representation decouples the number of 3D Gaussians from the number and resolution of input images, enabling training with many high-resolution input views. We train and evaluate our model on an internal dataset with more than 10,000 subjects, which is an order of magnitude larger than existing multi-view human head datasets. HeadsUp achieves state-of-the-art reconstruction quality and generalizes to novel identities without test-time optimization. We extensively analyze the scaling behavior of our model across identities, views, and model capacity, revealing practical insights for quality-compute trade-offs. Finally, we highlight the strength of our latent space by showcasing two downstream applications: generating novel 3D identities and animating the 3D heads with expression blendshapes.
title Large-Scale High-Quality 3D Gaussian Head Reconstruction from Multi-View Captures
topic Computer Vision and Pattern Recognition
Machine Learning
url https://arxiv.org/abs/2605.04035