SwiftVGGT: A Scalable Visual Geometry Grounded Transformer for Large-Scale Scenes

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Lee, Jungho, Lee, Minhyeok, Yang, Sunghun, Kang, Minseok, Lee, Sangyoun
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866908670760058880
author Lee, Jungho
Lee, Minhyeok
Yang, Sunghun
Kang, Minseok
Lee, Sangyoun
author_facet Lee, Jungho
Lee, Minhyeok
Yang, Sunghun
Kang, Minseok
Lee, Sangyoun
contents 3D reconstruction in large-scale scenes is a fundamental task in 3D perception, but the inherent trade-off between accuracy and computational efficiency remains a significant challenge. Existing methods either prioritize speed and produce low-quality results, or achieve high-quality reconstruction at the cost of slow inference times. In this paper, we propose SwiftVGGT, a training-free method that significantly reduce inference time while preserving high-quality dense 3D reconstruction. To maintain global consistency in large-scale scenes, SwiftVGGT performs loop closure without relying on the external Visual Place Recognition (VPR) model. This removes redundant computation and enables accurate reconstruction over kilometer-scale environments. Furthermore, we propose a simple yet effective point sampling method to align neighboring chunks using a single Sim(3)-based Singular Value Decomposition (SVD) step. This eliminates the need for the Iteratively Reweighted Least Squares (IRLS) optimization commonly used in prior work, leading to substantial speed-ups. We evaluate SwiftVGGT on multiple datasets and show that it achieves state-of-the-art reconstruction quality while requiring only 33% of the inference time of recent VGGT-based large-scale reconstruction approaches.
format Preprint
id arxiv_https___arxiv_org_abs_2511_18290
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle SwiftVGGT: A Scalable Visual Geometry Grounded Transformer for Large-Scale Scenes
Lee, Jungho
Lee, Minhyeok
Yang, Sunghun
Kang, Minseok
Lee, Sangyoun
Computer Vision and Pattern Recognition
Artificial Intelligence
3D reconstruction in large-scale scenes is a fundamental task in 3D perception, but the inherent trade-off between accuracy and computational efficiency remains a significant challenge. Existing methods either prioritize speed and produce low-quality results, or achieve high-quality reconstruction at the cost of slow inference times. In this paper, we propose SwiftVGGT, a training-free method that significantly reduce inference time while preserving high-quality dense 3D reconstruction. To maintain global consistency in large-scale scenes, SwiftVGGT performs loop closure without relying on the external Visual Place Recognition (VPR) model. This removes redundant computation and enables accurate reconstruction over kilometer-scale environments. Furthermore, we propose a simple yet effective point sampling method to align neighboring chunks using a single Sim(3)-based Singular Value Decomposition (SVD) step. This eliminates the need for the Iteratively Reweighted Least Squares (IRLS) optimization commonly used in prior work, leading to substantial speed-ups. We evaluate SwiftVGGT on multiple datasets and show that it achieves state-of-the-art reconstruction quality while requiring only 33% of the inference time of recent VGGT-based large-scale reconstruction approaches.
title SwiftVGGT: A Scalable Visual Geometry Grounded Transformer for Large-Scale Scenes
topic Computer Vision and Pattern Recognition
Artificial Intelligence
url https://arxiv.org/abs/2511.18290