Adaptive World Memory 3D Foundation Model for Scalable 3D Mapping, Localization, and Rendering

原文：Adaptive World Memory 3D Foundation Model for Scalable 3D Mapping, Localization, and Rendering

作者：Tianchen Deng, Guole Shen, Yilin Shen, Wenhua Wu, Yilin Fang, Ziqi Ma, Tianjun Zhang, Shenghai Yuan, Wolfram Burgard, Hesheng Wang

来源：arXiv cs.CV（计算机视觉）

正文

Computer Science > Computer Vision and Pattern Recognition

arXiv:2609.21502v1 (cs)

[Submitted on 18 Sep 2026]

Title:Adaptive World Memory 3D Foundation Model for Scalable 3D Mapping, Localization, and Rendering

Authors:Tianchen Deng, Guole Shen, Yilin Shen, Wenhua Wu, Yilin Fang, Ziqi Ma, Tianjun Zhang, Shenghai Yuan, Wolfram Burgard, Hesheng Wang

View a PDF of the paper titled Adaptive World Memory 3D Foundation Model for Scalable 3D Mapping, Localization, and Rendering, by Tianchen Deng and 9 other authors

View PDF

HTML (experimental)

Abstract:Recent 3D foundation models enable generalizable geometric reasoning from RGB images but remain limited in persistent memory, scalability, and renderable scene modeling. We present a memory-centric 3D foundation model for scalable robotic localization, reconstruction, and Gaussian rendering. Its core is an adaptive world memory mechanism that combines transformer-based gated updates with test-time temporal-spatial regulation. Learned gates control recurrent memory propagation, while temporal state evolution and spatial observation-state consistency regulate token-wise updates and forgetting over long image sequences. To support large-scale mapping, we organize memory into local submaps and integrate progressive mapping and tracking, loop closure, and SL(4)-based global refinement to maintain local accuracy and global consistency. A Gaussian reconstruction head decodes memory-enhanced features into renderable primitives, unifying camera pose estimation, dense point-cloud reconstruction, and photorealistic rendering within a single model. Experiments on public benchmarks and self-collected datasets from diverse robotic platforms demonstrate improved trajectory accuracy, reconstruction completeness, and rendering quality over existing 3D foundation reconstruction and SLAM baselines. These results support adaptive memory as a foundation for persistent robotic world modeling. The dataset and code will be made publicly available at \href{this https URL}{this https URL}.

Subjects:

Computer Vision and Pattern Recognition (cs.CV); Robotics (cs.RO)

Cite as:

arXiv:2609.21502 [cs.CV]

(or

arXiv:2609.21502v1 [cs.CV] for this version)

https://doi.org/10.48550/arXiv.2609.21502

Focus to learn more

arXiv-issued DOI via DataCite (pending registration)

主题

计算机视觉 · 机器人

由「前沿雷达」于 2026-09-18 采集。正文取自原文页面，已保留出处链接。标题与正文版权归原作者所有。