Capture, Canonicalize, Splat: Zero-Shot 3D Gaussian Avatars from Unstructured Phone Images

Fuente: arXiv
Enregistré dans:
Détails bibliographiques
Auteurs principaux: Garbin, Emanuel, Adam, Guy, Krams, Oded, Barzelay, Zohar, Guendelman, Eran, Schwarz, Michael, Presutto, Matteo, Vatelmacher, Moran, Shenkman, Yigal, Peker, Eli, Druker, Itai, Patish, Uri, Blum, Yoav, Bluvstein, Max, Li, Junxuan, Khirodkar, Rawal, Saito, Shunsuke
Format: Preprint
Publié: 2025
Sujets:
Accès en ligne:
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866917044448919552
author Garbin, Emanuel
Adam, Guy
Krams, Oded
Barzelay, Zohar
Guendelman, Eran
Schwarz, Michael
Presutto, Matteo
Vatelmacher, Moran
Shenkman, Yigal
Peker, Eli
Druker, Itai
Patish, Uri
Blum, Yoav
Bluvstein, Max
Li, Junxuan
Khirodkar, Rawal
Saito, Shunsuke
author_facet Garbin, Emanuel
Adam, Guy
Krams, Oded
Barzelay, Zohar
Guendelman, Eran
Schwarz, Michael
Presutto, Matteo
Vatelmacher, Moran
Shenkman, Yigal
Peker, Eli
Druker, Itai
Patish, Uri
Blum, Yoav
Bluvstein, Max
Li, Junxuan
Khirodkar, Rawal
Saito, Shunsuke
contents We present a novel, zero-shot pipeline for creating hyperrealistic, identity-preserving 3D avatars from a few unstructured phone images. Existing methods face several challenges: single-view approaches suffer from geometric inconsistencies and hallucinations, degrading identity preservation, while models trained on synthetic data fail to capture high-frequency details like skin wrinkles and fine hair, limiting realism. Our method introduces two key contributions: (1) a generative canonicalization module that processes multiple unstructured views into a standardized, consistent representation, and (2) a transformer-based model trained on a new, large-scale dataset of high-fidelity Gaussian splatting avatars derived from dome captures of real people. This "Capture, Canonicalize, Splat" pipeline produces static quarter-body avatars with compelling realism and robust identity preservation from unstructured photos.
format Preprint
id arxiv_https___arxiv_org_abs_2510_14081
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Capture, Canonicalize, Splat: Zero-Shot 3D Gaussian Avatars from Unstructured Phone Images
Garbin, Emanuel
Adam, Guy
Krams, Oded
Barzelay, Zohar
Guendelman, Eran
Schwarz, Michael
Presutto, Matteo
Vatelmacher, Moran
Shenkman, Yigal
Peker, Eli
Druker, Itai
Patish, Uri
Blum, Yoav
Bluvstein, Max
Li, Junxuan
Khirodkar, Rawal
Saito, Shunsuke
Computer Vision and Pattern Recognition
Graphics
We present a novel, zero-shot pipeline for creating hyperrealistic, identity-preserving 3D avatars from a few unstructured phone images. Existing methods face several challenges: single-view approaches suffer from geometric inconsistencies and hallucinations, degrading identity preservation, while models trained on synthetic data fail to capture high-frequency details like skin wrinkles and fine hair, limiting realism. Our method introduces two key contributions: (1) a generative canonicalization module that processes multiple unstructured views into a standardized, consistent representation, and (2) a transformer-based model trained on a new, large-scale dataset of high-fidelity Gaussian splatting avatars derived from dome captures of real people. This "Capture, Canonicalize, Splat" pipeline produces static quarter-body avatars with compelling realism and robust identity preservation from unstructured photos.
title Capture, Canonicalize, Splat: Zero-Shot 3D Gaussian Avatars from Unstructured Phone Images
topic Computer Vision and Pattern Recognition
Graphics
url https://arxiv.org/abs/2510.14081