---
格式版本: 2
标题: "GS-CPE: Unified 6-Degree-of-Freedom Camera Pose Estimation via 3D Gaussian Splatting"
原文链接: "https://arxiv.org/abs/2608.10938"
发布日期: "2026-08-11"
发布时间校准状态: "found"
发布时间需复核: "否"
发布时间来源: "rule:local:strict_original_body"
发布时间证据: "**\\[v1\\]** Tue, 11 Aug 2026 14:06:51 UTC (1,496 KB)"
发布时间校准原因: "规则确认唯一严格发布时间，来源 local:strict_original_body"
发布时间校准置信度: "high"
发布时间候选数量: 18
发布时间严格候选数量: 6
发布时间原页读取状态: "source template page reused from URL open"
发布时间未找到原因: ""
发布时间校准时间: "2026-08-12T11:21:48+08:00"
发布时间仲裁状态: "skipped"
发布时间仲裁尝试次数: 0
发布时间仲裁耗时毫秒: 0
发现时间: "2026-08-12T11:16:29+08:00"
入库时间: "2026-08-12T03:21:48.923Z"
来源平台: "arXiv 学术论文搜索"
搜索渠道: "source_template"
搜索词: "https://arxiv.org/search/?query=performance&searchtype=all"
匹配关键词:
  - "performance"
相关厂家:
  []
相关专家:
  []
内容类型: "网页"
抓取工具: "Free Fetch + Defuddle"
清洗工具: "Defuddle Markdown + Defuddle/Readability 正文提取"
原始附件:
  []
AI优质: "否"
AI打分: 17
AI分档: "非优质"
AI质检状态: "不通过"
AI打分理由: "该资料为计算机视觉领域论文，与超节点/AI Rack/机柜级AI基础设施完全无关。"
AI质检模型: "ali-deepseek-v4-flash"
AI质检时间: "2026-08-12T11:21:56+08:00"
AI主题相关性: 0
AI来源权威性: 7
AI新颖性: 5
AI技术细节: 0
AI商业部署信号: 0
AI完整性: 5
采集批次: "2026年8月12日2点04分20秒"
采集批次ID: "20260812-020420-289"
去重键: "https://arxiv.org/abs/2608.10938"
---

## Computer Science > Computer Vision and Pattern Recognition

## Title:GS-CPE: Unified 6-Degree-of-Freedom Camera Pose Estimation via 3D Gaussian Splatting

Authors:[Huaiyuan Weng](https://arxiv.org/search/cs?searchtype=author&query=Weng,+H), [Chul Min Yeum](https://arxiv.org/search/cs?searchtype=author&query=Yeum,+C+M), [Su-Min Kang](https://arxiv.org/search/cs?searchtype=author&query=Kang,+S)

[View PDF](https://arxiv.org/pdf/2608.10938) [HTML (experimental)](https://arxiv.org/html/2608.10938v1)

> Abstract:Despite substantial progress in visual localization, from scene coordinate regression to direct camera pose regression, achieving both robust generalization and high accuracy remain challenging. This study introduces GS-CPE (Gaussian Splatting based Camera Pose Estimation), a coarse-to-fine framework for 6-DoF camera pose estimation that unifies geometry-based coarse pose estimation with robust 3D Gaussian Splatting (3DGS) warping based pose refinement. GS-CPE first estimates a coarse pose via retrieval-guided geometric pose estimation on a 3DGS scene representation, then refines it by minimizing a visibility aware masked RGB warping objective in a multi-scale optimization framework, with adaptive re-rendering. Extensive experiments on indoor and outdoor benchmarks including 7Scenes, Cambridge Landmarks, FAST-LIVO2 datasets, and a custom dataset demonstrate state-of-the-art performance, consistently outperforming in both accuracy and generalization.

| Comments: |  |
| --- | --- |
| Subjects: | Computer Vision and Pattern Recognition (cs.CV) |
| Cite as: | [arXiv:2608.10938](https://arxiv.org/abs/2608.10938) \[cs.CV\] |
|  | (or [arXiv:2608.10938v1](https://arxiv.org/abs/2608.10938v1) \[cs.CV\] for this version) |
|  | [https://doi.org/10.48550/arXiv.2608.10938](https://doi.org/10.48550/arXiv.2608.10938) |

## Submission history

From: Huaiyuan Weng \[[view email](https://arxiv.org/show-email/ec0deb4e/2608.10938)\]  
**\[v1\]** Tue, 11 Aug 2026 14:06:51 UTC (1,496 KB)

[Which authors of this paper are endorsers?](https://arxiv.org/auth/show-endorsers/2608.10938) | Disable MathJax ([What is MathJax?](https://info.arxiv.org/help/mathjax.html))