InstantSfM: Towards GPU-Native SfM for the Deep Learning Era

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Zhong, Jiankun, Zhan, Zitong, Gao, Quankai, Chen, Ziyu, Lou, Haozhe, Mao, Jiageng, Neumann, Ulrich, Wang, Chen, Wang, Yue
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912960942702592
author Zhong, Jiankun
Zhan, Zitong
Gao, Quankai
Chen, Ziyu
Lou, Haozhe
Mao, Jiageng
Neumann, Ulrich
Wang, Chen
Wang, Yue
author_facet Zhong, Jiankun
Zhan, Zitong
Gao, Quankai
Chen, Ziyu
Lou, Haozhe
Mao, Jiageng
Neumann, Ulrich
Wang, Chen
Wang, Yue
contents Structure-from-Motion (SfM) is a fundamental technique for recovering camera poses and scene structure from multi-view imagery, serving as a critical upstream component for applications ranging from 3D reconstruction to modern neural scene representations such as 3D Gaussian Splatting. However, most mature SfM systems remain CPU-centric and built upon traditional optimization toolchains, creating a growing mismatch with modern GPU-based, learning-driven pipelines and limiting scalability in large-scale scenes. While recent advances in GPU-accelerated bundle adjustment (BA) have demonstrated the potential of parallel sparse optimization, extending these techniques to build a complete global SfM system remains challenging due to unresolved issues in metric scale recovery and numerical robustness. In this paper, we implement a fully GPU-based and PyTorch-compatible global SfM system, named InstantSfM, to integrate seamlessly with modern learning pipelines. InstantSfM embeds metric depth priors directly into both global positioning and BA through a depth-constrained Jacobian structure, thereby resolving scale ambiguity within the optimization framework. To ensure numerical stability, we employ explicit filtering of under-constrained variables for the Jacobian matrix in an optimized GPU-friendly manner. Extensive experiments on diverse datasets demonstrate that InstantSfM achieves state-of-the-art efficiency while maintaining reconstruction accuracy comparable to both established classical pipelines and recent learning-based methods, showing up to ${\sim40\times}$ speedup over COLMAP on large-scale scenes.
format Preprint
id arxiv_https___arxiv_org_abs_2510_13310
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle InstantSfM: Towards GPU-Native SfM for the Deep Learning Era
Zhong, Jiankun
Zhan, Zitong
Gao, Quankai
Chen, Ziyu
Lou, Haozhe
Mao, Jiageng
Neumann, Ulrich
Wang, Chen
Wang, Yue
Computer Vision and Pattern Recognition
Structure-from-Motion (SfM) is a fundamental technique for recovering camera poses and scene structure from multi-view imagery, serving as a critical upstream component for applications ranging from 3D reconstruction to modern neural scene representations such as 3D Gaussian Splatting. However, most mature SfM systems remain CPU-centric and built upon traditional optimization toolchains, creating a growing mismatch with modern GPU-based, learning-driven pipelines and limiting scalability in large-scale scenes. While recent advances in GPU-accelerated bundle adjustment (BA) have demonstrated the potential of parallel sparse optimization, extending these techniques to build a complete global SfM system remains challenging due to unresolved issues in metric scale recovery and numerical robustness. In this paper, we implement a fully GPU-based and PyTorch-compatible global SfM system, named InstantSfM, to integrate seamlessly with modern learning pipelines. InstantSfM embeds metric depth priors directly into both global positioning and BA through a depth-constrained Jacobian structure, thereby resolving scale ambiguity within the optimization framework. To ensure numerical stability, we employ explicit filtering of under-constrained variables for the Jacobian matrix in an optimized GPU-friendly manner. Extensive experiments on diverse datasets demonstrate that InstantSfM achieves state-of-the-art efficiency while maintaining reconstruction accuracy comparable to both established classical pipelines and recent learning-based methods, showing up to ${\sim40\times}$ speedup over COLMAP on large-scale scenes.
title InstantSfM: Towards GPU-Native SfM for the Deep Learning Era
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2510.13310