# MuDG: Taming Multi-modal Diffusion with Gaussian Splatting for Urban Scene Reconstruction

arXiv: https://arxiv.org/abs/2503.10604 · HTML: https://mlsanity.com/paper/2503.10604/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- Placed in World models for autonomous driving, one of 11 directions in World models.
- Builds on 44 of its 49 references that are in the map.
- First posted on arXiv 13 Mar 2025.

## Placed in

- World models / [World models for autonomous driving](https://mlsanity.com/wm/world-models-for-autonomous-driving/)
- Diffusion models / [3D, point cloud and scene generation](https://mlsanity.com/diffusion/3d-point-cloud-and-scene-generation/)

## Builds on (most cited first, 30 of 44)

- [Denoising Diffusion Probabilistic Models](https://mlsanity.com/paper/2006.11239/) (2020)
- [Denoising Diffusion Implicit Models](https://mlsanity.com/paper/2010.02502/) (2020)
- [3D Gaussian Splatting for Real-Time Radiance Field Rendering](https://mlsanity.com/paper/2308.04079/) (2023)
- [SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers](https://mlsanity.com/paper/2105.15203/) (2021)
- [Stable Video Diffusion: Scaling Latent Video Diffusion Models to Large Datasets](https://mlsanity.com/paper/2311.15127/) (2023)
- [Scalability in Perception for Autonomous Driving: Waymo Open Dataset](https://mlsanity.com/paper/1912.04838/) (2019)
- [Plenoxels: Radiance Fields without Neural Networks](https://mlsanity.com/paper/2112.05131/) (2021)
- [2D Gaussian Splatting for Geometrically Accurate Radiance Fields](https://mlsanity.com/paper/2403.17888/) (2024)
- [NeRF++: Analyzing and Improving Neural Radiance Fields](https://mlsanity.com/paper/2010.07492/) (2020)
- [Wonder3D: Single Image to 3D using Cross-Domain Diffusion](https://mlsanity.com/paper/2310.15008/) (2023)
- [Mip-Splatting: Alias-free 3D Gaussian Splatting](https://mlsanity.com/paper/2311.16493/) (2023)
- [DynamiCrafter: Animating Open-domain Images with Video Diffusion Priors](https://mlsanity.com/paper/2310.12190/) (2023)
- [ViewCrafter: Taming Video Diffusion Models for High-fidelity Novel View Synthesis](https://mlsanity.com/paper/2409.02048/) (2024)
- [Video-to-Video Synthesis](https://mlsanity.com/paper/1808.06601/) (2018)
- [DrivingGaussian: Composite Gaussian Splatting for Surrounding Dynamic Autonomous Driving Scenes](https://mlsanity.com/paper/2312.07920/) (2023)
- [Vista: A Generalizable Driving World Model with High Fidelity and Versatile Controllability](https://mlsanity.com/paper/2405.17398/) (2024)
- [Driving into the Future: Multiview Visual Forecasting and Planning with World Model for Autonomous Driving](https://mlsanity.com/paper/2311.17918/) (2023)
- [UniSim: A Neural Closed-Loop Sensor Simulator](https://mlsanity.com/paper/2308.01898/) (2023)
- [GeoWizard: Unleashing the Diffusion Priors for 3D Geometry Estimation from a Single Image](https://mlsanity.com/paper/2403.12013/) (2024)
- [DriveDreamer: Towards Real-world-driven World Models for Autonomous Driving](https://mlsanity.com/paper/2309.09777/) (2023)
- [MagicDrive: Street View Generation with Diverse 3D Geometry Control](https://mlsanity.com/paper/2310.02601/) (2023)
- [Street Gaussians: Modeling Dynamic Urban Scenes with Gaussian Splatting](https://mlsanity.com/paper/2401.01339/) (2024)
- [EmerNeRF: Emergent Spatial-Temporal Scene Decomposition via Self-Supervision](https://mlsanity.com/paper/2311.02077/) (2023)
- [GaussianPro: 3D Gaussian Splatting with Progressive Propagation](https://mlsanity.com/paper/2402.14650/) (2024)
- [DriveDreamer-2: LLM-Enhanced World Models for Diverse Driving Video Generation](https://mlsanity.com/paper/2403.06845/) (2024)
- [NeuRAD: Neural Rendering for Autonomous Driving](https://mlsanity.com/paper/2311.15260/) (2023)
- [Periodic Vibration Gaussian: Dynamic Urban Scene Reconstruction and Real-time Rendering](https://mlsanity.com/paper/2311.18561/) (2023)
- [HUGS: Holistic Urban 3D Scene Understanding via Gaussian Splatting](https://mlsanity.com/paper/2403.12722/) (2024)
- [DriveDreamer4D: World Models Are Effective Data Machines for 4D Driving Scene Representation](https://mlsanity.com/paper/2410.13571/) (2024)
- [UniScene: Unified Occupancy-centric Driving Scene Generation](https://mlsanity.com/paper/2412.05435/) (2024)

## Built on by (most cited first, 8 of 8)

- [3D and 4D World Modeling: A Survey](https://mlsanity.com/paper/2509.07996/) (2025)
- [GaussianDWM: 3D Gaussian Driving World Model for Unified Scene Understanding and Multi-Modal Generation](https://mlsanity.com/paper/2512.23180/) (2025)
- [RoDyn: Taming Interactive Robot-Dynamic 2.5D World Model for Robotic Manipulation](https://mlsanity.com/paper/2510.09036/) (2025)
- [OmniGen: Unified Multimodal Sensor Generation for Autonomous Driving](https://mlsanity.com/paper/2512.14225/) (2025)
- [Diffusion-guided Generalizable Enhancer for Urban Scene Reconstruction](https://mlsanity.com/paper/2605.22420/) (2026)
- [DriveExplorer: Images-Only Decoupled 4D Reconstruction with Progressive Restoration for Driving View Extrapolation](https://mlsanity.com/paper/2512.23983/) (2025)
- [UniPR: Unified Object-level Real-to-Sim Perception and Reconstruction from a Single Stereo Pair](https://mlsanity.com/paper/2603.19616/) (2026)
- [GaussianDWM++: Language-Grounded 3D Gaussian Driving World Model for Unified Scene Understanding, Editing, and Multi-Modal Generation](https://mlsanity.com/paper/2608.16234/) (2026)

## Abstract (arXiv)

Recent breakthroughs in radiance fields have significantly advanced 3D scene reconstruction and novel view synthesis (NVS) in autonomous driving. Nevertheless, critical limitations persist: reconstruction-based methods exhibit substantial performance deterioration under significant viewpoint deviations from training trajectories, while generation-based techniques struggle with temporal coherence and precise scene controllability. To overcome these challenges, we present MuDG, an innovative framework that integrates Multi-modal Diffusion model with Gaussian Splatting (GS) for Urban Scene Reconstruction. MuDG leverages aggregated LiDAR point clouds with RGB and geometric priors to condition a multi-modal video diffusion model, synthesizing photorealistic RGB, depth, and semantic outputs for novel viewpoints. This synthesis pipeline enables feed-forward NVS without computationally intensive per-scene optimization, providing comprehensive supervision signals to refine 3DGS representations for rendering robustness enhancement under extreme viewpoint changes. Experiments on the Open Waymo Dataset demonstrate that MuDG outperforms existing methods in both reconstruction and synthesis quality.