World SLAM Model: Joint World Modeling for SLAM and Navigation
cs.RO, cs.CV
Submitted: 2026-09-26
Updated: 2026-09-26
Comments: Website: https://tsinghua-mars-lab.github.io/WorldSLAMModel
Project page: https://tsinghua-mars-lab.github.io/WorldSLAMModel
License: http://creativecommons.org/licenses/by/4.0/
The gist: We introduce World SLAM Model (WSM), a unified framework that brings the SLAM paradigm directly into downstream navigation.
Terminology
Abstract
We introduce World SLAM Model (WSM), a unified framework that brings the SLAM paradigm directly into downstream navigation. Rather than treating SLAM merely as an upstream module that provides poses, maps or tokens, WSM adopts its core mechanisms, including incremental state updates with persistent memory and backend refinement of accumulated errors, to maintain a consistent world state during interaction. Given the current observation and a navigation goal, WSM predicts future visual states and jointly estimates their camera motion and dense geometry, grounding visual prediction in an evolving spatial world state. This spatial state is continuously updated as new observations arrive and provides the basis for action generation and closed-loop navigation. WSM is trained end-to-end with a joint navigation--SLAM objective, enabling downstream navigation to benefit directly from SLAM-style state maintenance and refinement while preserving accurate geometric estimation. Experiments demonstrate improved navigation performance together with strong SLAM accuracy, highlighting the potential of SLAM as an intrinsic mechanism for long-horizon world modeling and embodied interaction.
Sources
- Cosmos 3: Omnimodal World Models for Physical AI
- NavDP: Learning Sim-to-Real Navigation Diffusion Policy with Privileged Information Guidance
- Learning to Explore using Active Neural SLAM
- Geometric Context Transformer for Streaming 3D Reconstruction
- LoRA: Low-Rank Adaptation of Large Language Models
- Causal World Modeling for Robot Control
- Flow Matching for Generative Modeling
- LoGoPlanner: Localization Grounded Navigation Policy with Metric-aware Visual Geometry
- ViNT: A Foundation Model for Visual Navigation
- Wan: Open and Advanced Large-Scale Video Generative Models
- StreamVLN: Streaming Vision-and-Language Navigation via SlowFast Context Modeling
- DD-PPO: Learning Near-Perfect PointGoal Navigators from 2.5 Billion Frames
- iPlanner: Imperative Path Planning
- Fast-WAM: Do World Action Models Need Test-time Future Imagination?
- JanusVLN: Decoupling Semantics and Spatiality with Dual Implicit Memory for Vision-Language Navigation
Related papers
- FMT x: An Efficient and Asymptotically Optimal Extension of the Fast Marching Tree for Dynamic Replanning
- MPCFormer: A physics-informed data-driven approach for explainable socially-aware autonomous driving
- RoboLab: A High-Fidelity Simulation Benchmark for Analysis of Task Generalist Policies
- ARC: A Reasoning Recipe for Robot Foundation Models
- HRDexDB: A 4D Dexterous Grasping Dataset Across Human and Multiple Robot Embodiments
- APT: Action Expert Pretraining Improves Instruction Generalization of Vision-Language-Action Policies