GeoState: Learning View-Decoupled Geometry States for 3D World Modeling
Summary
GeoState提出了一种视角解耦的几何表示，将场景动态与相机运动分离以实现预测性3D世界建模，在Waymo上超越先前方法，并可零样本泛化到KITTI和nuScenes。
View Cached Full Text
Cached at: 10/19/26, 03:26 AM
Paper page - GeoState: Learning View-Decoupled Geometry States for 3D World Modeling
Source: https://huggingface.co/papers/2610.22343
Abstract
GeoState introduces a view-independent geometric representation that separates scene dynamics from camera motion for predictive 3D world modeling, enabling consistent multi-view forecasting and cross-dataset generalization.
Recent 3D world models increasingly use features fromgeometry foundation models(GFMs) as predictive world states, benefiting from their rich geometric priors. However, the view dependence of these features entangles scene dynamics with camera-induced variation, making the resulting state evolution harder to predict. We propose GeoState, aview-decoupled geometry representationfor predictive 3D world modeling that separates scene-state evolution from camera-dependent observation. Specifically, ascene tokenizeraggregates multi-view GFM features into a view-decoupled scene state, which ageometry decoderqueries usingcamera rays. Anautoregressive world modelthen predicts future scene states, while camera motion is modeled separately. The predicted states can therefore be decoded under either predicted or externally specified cameras, while retaining a single context-derived scale throughout the rollout without per-frame scale alignment. Experiments show that GeoState outperforms prior 3D world models on Waymo, supports geometrically consistentcross-view querying, and generalizes zero-shot to KITTI and nuScenes. Project page: https://hongchang-liu.github.io/GeoState-project-page/
View arXiv pageView PDFProject pageGitHub9Add to collection
Models citing this paper0
No model linking this paper
Cite arxiv.org/abs/2610.22343 in a model README.md to link it from this page.
Datasets citing this paper0
No dataset linking this paper
Cite arxiv.org/abs/2610.22343 in a dataset README.md to link it from this page.
Spaces citing this paper0
No Space linking this paper
Cite arxiv.org/abs/2610.22343 in a Space README.md to link it from this page.
Collections including this paper0
No Collection including this paper
Add this paper to acollectionto link it from this page.
Similar Articles
Drive3R: Block-wise Streaming Geometric Foundation Models for Autonomous Driving
Drive3R 提出一种块状流式几何基础模型，通过块状因果注意力和置信度感知重锚定实现可扩展的长时程多视图 3D 重建，在 nuScenes、KITTI 和 Waymo 上达到新的最先进水平。
GeoVerse: World-Consistent Novel View Synthesis in Geometric Latent Space
GeoVerse is a framework for synthesizing world-consistent novel views by performing generation in a geometric latent space from a 3D foundation model and injecting appearance priors from a video generative model, demonstrating improved visual quality and geometric consistency over existing methods.
SG-WAM: Self-Guided World Modeling in Geometry-Aware Policy Space
This paper proposes SG-WAM, a self-guided framework for learning geometry-aware action-conditioned world models directly in policy-derived representation space. It achieves state-of-the-art success rates on LIBERO and LIBERO-Plus benchmarks, outperforming strong baselines in real-world evaluations.
GAE: Learning a Geometry-Native Latent Space for 3D-Consistent World Generation
The paper introduces GAE, a geometry-native autoencoder that creates a compact latent space for generating 3D-consistent scenes, enhancing visual quality and coherence over existing methods.
WorldRecon: Generative Multi-View Mesh Reconstruction for Large-Scale 3D World
WorldRecon 提出一种生成式多视角网格重建框架，可从无位姿 RGB 图像重建大规模 3D 场景的完整网格。它将 TRELLIS.2 生成模型与重建式条件架构结合，通过邻域感知潜在修补和因果世界生成保证场景级一致性，在合成与真实室内场景上分别将 Chamfer 距离降低 14.9% 和 43.1%，达到最先进性能。