# STORM: Spatio-Temporal Reconstruction Model for Large-Scale Outdoor Scenes

arXiv: https://arxiv.org/abs/2501.00602 · HTML: https://mlsanity.com/paper/2501.00602/

## Key facts

- Cited by 55 papers in the map (data through 14 Oct 2026).
- Builds on 72 of its 78 references that are in the map.
- First posted on arXiv 31 Dec 2024.

## Builds on (most cited first, 30 of 72)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://mlsanity.com/paper/2010.11929/) (2020)
- [High-Resolution Image Synthesis with Latent Diffusion Models](https://mlsanity.com/paper/2112.10752/) (2021)
- [Very Deep Convolutional Networks for Large-Scale Image Recognition](https://mlsanity.com/paper/1409.1556/) (2014)
- [The Unreasonable Effectiveness of Deep Features as a Perceptual Metric](https://mlsanity.com/paper/1801.03924/) (2018)
- [Segment Anything](https://mlsanity.com/paper/2304.02643/) (2023)
- [3D Gaussian Splatting for Real-Time Radiance Field Rendering](https://mlsanity.com/paper/2308.04079/) (2023)
- [Scalable Diffusion Models with Transformers](https://mlsanity.com/paper/2212.09748/) (2022)
- [nuScenes: A multimodal dataset for autonomous driving](https://mlsanity.com/paper/1903.11027/) (2019)
- [ShapeNet: An Information-Rich 3D Model Repository](https://mlsanity.com/paper/1512.03012/) (2015)
- [Scalability in Perception for Autonomous Driving: Waymo Open Dataset](https://mlsanity.com/paper/1912.04838/) (2019)
- [pixelNeRF: Neural Radiance Fields from One or Few Images](https://mlsanity.com/paper/2012.02190/) (2020)
- [Objaverse: A Universe of Annotated 3D Objects](https://mlsanity.com/paper/2212.08051/) (2022)
- [D-NeRF: Neural Radiance Fields for Dynamic Scenes](https://mlsanity.com/paper/2011.13961/) (2020)
- [4D Gaussian Splatting for Real-Time Dynamic Scene Rendering](https://mlsanity.com/paper/2310.08528/) (2023)
- [Scaling Vision Transformers](https://mlsanity.com/paper/2106.04560/) (2021)
- [Objaverse-XL: A Universe of 10M+ 3D Objects](https://mlsanity.com/paper/2307.05663/) (2023)
- [Deformable 3D Gaussians for High-Fidelity Monocular Dynamic Scene Reconstruction](https://mlsanity.com/paper/2309.13101/) (2023)
- [Neural Scene Flow Fields for Space-Time View Synthesis of Dynamic Scenes](https://mlsanity.com/paper/2011.13084/) (2020)
- [K-Planes: Explicit Radiance Fields in Space, Time, and Appearance](https://mlsanity.com/paper/2301.10241/) (2023)
- [LGM: Large Multi-View Gaussian Model for High-Resolution 3D Content Creation](https://mlsanity.com/paper/2402.05054/) (2024)
- [Dynamic 3D Gaussians: Tracking by Persistent Dynamic View Synthesis](https://mlsanity.com/paper/2308.09713/) (2023)
- [Argoverse 2: Next Generation Datasets for Self-Driving Perception and Forecasting](https://mlsanity.com/paper/2301.00493/) (2023)
- [IBRNet: Learning Multi-View Image-Based Rendering](https://mlsanity.com/paper/2102.13090/) (2021)
- [pixelSplat: 3D Gaussian Splats from Image Pairs for Scalable Generalizable 3D Reconstruction](https://mlsanity.com/paper/2312.12337/) (2023)
- [Common Objects in 3D: Large-Scale Learning and Evaluation of Real-life 3D Category Reconstruction](https://mlsanity.com/paper/2109.00512/) (2021)
- [HexPlane: A Fast Representation for Dynamic Scenes](https://mlsanity.com/paper/2301.09632/) (2023)
- [Vision Transformer Adapter for Dense Predictions](https://mlsanity.com/paper/2205.08534/) (2022)
- [CoTracker: It is Better to Track Together](https://mlsanity.com/paper/2307.07635/) (2023)
- [Real-time Photorealistic Dynamic Scene Representation and Rendering with 4D Gaussian Splatting](https://mlsanity.com/paper/2310.10642/) (2023)

## Built on by (most cited first, 30 of 55)

- [Pseudo-Simulation for Autonomous Driving](https://mlsanity.com/paper/2506.04218/) (2025)
- [4DGT: Learning a 4D Gaussian Transformer Using Real-World Monocular Videos](https://mlsanity.com/paper/2506.08015/) (2025)
- [DiST-4D: Disentangled Spatiotemporal Diffusion with Metric Depth for 4D Driving Scene Generation](https://mlsanity.com/paper/2503.15208/) (2025)
- [SimScale: Learning to Drive via Real-World Simulation at Scale](https://mlsanity.com/paper/2511.23369/) (2025)
- [MoVieS: Motion-Aware 4D Dynamic View Synthesis in One Second](https://mlsanity.com/paper/2507.10065/) (2025)
- [GaussianDWM: 3D Gaussian Driving World Model for Unified Scene Understanding and Multi-Modal Generation](https://mlsanity.com/paper/2512.23180/) (2025)
- [X-Scene: Large-Scale Driving Scene Generation with High Fidelity and Flexible Controllability](https://mlsanity.com/paper/2506.13558/) (2025)
- [MonoFusion: Sparse-View 4D Reconstruction via Monocular Fusion](https://mlsanity.com/paper/2507.23782/) (2025)
- [Unifying Appearance Codes and Bilateral Grids for Driving Scene Gaussian Splatting](https://mlsanity.com/paper/2506.05280/) (2025)
- [TokenGS: Decoupling 3D Gaussian Prediction from Pixels with Learnable Tokens](https://mlsanity.com/paper/2604.15239/) (2026)
- [4D-LRM: Large Space-Time Reconstruction Model From and To Any View at Any Time](https://mlsanity.com/paper/2506.18890/) (2025)
- [Reconstruct, Inpaint, Test-Time Finetune: Dynamic Novel-view Synthesis from Monocular Videos](https://mlsanity.com/paper/2507.12646/) (2025)
- [WorldSplat: Gaussian-Centric Feed-Forward 4D Scene Generation for Autonomous Driving](https://mlsanity.com/paper/2509.23402/) (2025)
- [Learning-based 3D Reconstruction in Autonomous Driving: A Comprehensive Survey](https://mlsanity.com/paper/2503.14537/) (2025)
- [MuDG: Taming Multi-modal Diffusion with Gaussian Splatting for Urban Scene Reconstruction](https://mlsanity.com/paper/2503.10604/) (2025)
- [StreamSplat: Towards Online Dynamic 3D Reconstruction from Uncalibrated Video Streams](https://mlsanity.com/paper/2506.08862/) (2025)
- [4D Driving Scene Generation With Stereo Forcing](https://mlsanity.com/paper/2509.20251/) (2025)
- [Generative AI for Autonomous Driving: Frontiers and Opportunities](https://mlsanity.com/paper/2505.08854/) (2025)
- [E3D-Bench: A Benchmark for End-to-End 3D Geometric Foundation Models](https://mlsanity.com/paper/2506.01933/) (2025)
- [Efficient Multi-Camera Tokenization with Triplanes for End-to-End Driving](https://mlsanity.com/paper/2506.12251/) (2025)
- [ReconDrive: Fast Feed-Forward 4D Gaussian Splatting for Autonomous Driving Scene Reconstruction](https://mlsanity.com/paper/2603.07552/) (2026)
- [Feed-Forward 3D Scene Modeling: A Problem-Driven Perspective](https://mlsanity.com/paper/2604.14025/) (2026)
- [B'ezierGS: Dynamic Urban Scene Reconstruction with B'ezier Curve Gaussian Splatting](https://mlsanity.com/paper/2506.22099/) (2025)
- [Xiaomi Auto World Model: A Joint World Model Integrating Reconstruction and Generation for Autonomous Driving](https://mlsanity.com/paper/2605.18137/) (2026)
- [SQS: Enhancing Sparse Perception Models via Query-based Splatting in Autonomous Driving](https://mlsanity.com/paper/2509.16588/) (2025)
- [IDSplat: Instance-Decomposed 3D Gaussian Splatting for Driving Scenes](https://mlsanity.com/paper/2511.19235/) (2025)
- [EVolSplat4D: Efficient Volume-based Gaussian Splatting for 4D Urban Scene Synthesis](https://mlsanity.com/paper/2601.15951/) (2026)
- [UFO: Unifying Feed-Forward and Optimization-based Methods for Large Driving Scene Modeling](https://mlsanity.com/paper/2602.20943/) (2026)
- [SLARM: Streaming and Language-Aligned Reconstruction Model for Dynamic Scenes](https://mlsanity.com/paper/2603.22893/) (2026)
- [Instant NuRec: Feed-Forward 3D Gaussian Reconstruction for Driving Scene Simulation](https://mlsanity.com/paper/2607.14203/) (2026)

## Abstract (arXiv)

We present STORM, a spatio-temporal reconstruction model designed for reconstructing dynamic outdoor scenes from sparse observations. Existing dynamic reconstruction methods often rely on per-scene optimization, dense observations across space and time, and strong motion supervision, resulting in lengthy optimization times, limited generalization to novel views or scenes, and degenerated quality caused by noisy pseudo-labels for dynamics. To address these challenges, STORM leverages a data-driven Transformer architecture that directly infers dynamic 3D scene representations--parameterized by 3D Gaussians and their velocities--in a single forward pass. Our key design is to aggregate 3D Gaussians from all frames using self-supervised scene flows, transforming them to the target timestep to enable complete (i.e., "amodal") reconstructions from arbitrary viewpoints at any moment in time. As an emergent property, STORM automatically captures dynamic instances and generates high-quality masks using only reconstruction losses. Extensive experiments on public datasets show that STORM achieves precise dynamic scene reconstruction, surpassing state-of-the-art per-scene optimization methods (+4.3 to 6.6 PSNR) and existing feed-forward approaches (+2.1 to 4.7 PSNR) in dynamic regions. STORM reconstructs large-scale outdoor scenes in 200ms, supports real-time rendering, and outperforms competitors in scene flow estimation, improving 3D EPE by 0.422m and Acc5 by 28.02%. Beyond reconstruction, we showcase four additional applications of our model, illustrating the potential of self-supervised learning for broader dynamic scene understanding.