OuroWorld: Bringing Any 3D World Alive as Diverse, Endlessly Looping 3D Cinemagraphs
Not on the map. Why?
Abstract (arXiv)
Recent 3D world models generate photorealistic, explorable scenes that remain frozen in time. OuroWorld is a mask-free framework that turns any static 3D Gaussian Splatting scene into a 3D cinemagraph: a dynamic scene with vivid, diverse motion looping seamlessly from any viewpoint. A vision-language model infers plausible dynamics and guides a video model to synthesize a reference video, which we lift and complete into multi-view videos. To learn from this imperfect supervision, we propose Inconsistency-Robust Periodic 4DGS: a Fourier-series deformation field guarantees looping by construction, while a Grounded Drift Field anchored at the reference view absorbs cross-view inconsistency. Unlike prior Eulerian methods limited to fluid-like motion, we capture general deformation, object motion, and illumination change. We introduce a ground-truth-free evaluation covering vividness, naturalness, loop seam coherence, and scene quality. On 39 reconstructed and generated scenes, OuroWorld outperforms all baselines and wins 70.8%-99.0% of user-study comparisons. Project page: https://ouroworld.userwei.com
Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 31 of its 141 references that are in the map. First posted on arXiv 8 Oct 2026.
Builds on
31 of 141 references.
References · 31
- 2025Wan: Open and Advanced Large-Scale Video Generative Modelscited by 2,536
- 2023Stable Video Diffusion: Scaling Latent Video Diffusion Models to Large Datasetscited by 2,526
- 2021SDEdit: Guided Image Synthesis and Editing with Stochastic Differential Equationscited by 1,903
- 2024HunyuanVideo: A Systematic Framework For Large Video Generative Modelscited by 1,597
- 2018Demystifying MMD GANscited by 1,268
- 2018Towards Accurate Generative Models of Video: A New Metric & Challengescited by 1,204
- 2025OpenAI GPT-5 System Cardcited by 910
- 2025Depth Anything 3: Recovering the Visual Space from Any Viewscited by 626
- 2023LucidDreamer: Domain-free Generation of 3D Gaussian Splatting Scenescited by 228
- 2021HyperNeRF: A Higher-Dimensional Representation for Topologically Varying Neural Radiance Fieldscited by 210
- 2018Stereo Magnification: Learning View Synthesis using Multiplane Imagescited by 206
- 2023Text-To-4D Dynamic Scene Generationcited by 204
19 more
- 2023DreamGaussian4D: Generative 4D Gaussian Splattingcited by 191
- 2025ViPE: Video Pose Engine for 3D Geometric Perceptioncited by 162
- 2024WildGaussians: 3D Gaussian Splatting in the Wildcited by 150
- 2026Seedance 2.0: Advancing Video Generation for World Complexitycited by 142
- 2024DimensionX: Create Any 3D and 4D Scenes from a Single Image with Controllable Video Diffusioncited by 139
- 2025HunyuanWorld 1.0: Generating Immersive, Explorable, and Interactive 3D Worlds from Words or Pixelscited by 91
- 2024Physics3D: Learning Physical Properties of 3D Gaussians via Video Diffusioncited by 76
- 2026Lyra 2.0: Explorable Generative 3D Worldscited by 36
- 2026HY-World 2.0: A Multi-Modal World Model for Reconstructing, Generating, and Simulating 3D Worldscited by 34
- 2024VividDream: Generating 3D Scene with Ambient Dynamicscited by 21
- 2026VGGT-Omegacited by 20
- 2019Animating Landscape: Self-Supervised Learning of Decoupled Motion and Appearance for Single-Image Video Synthesiscited by 15
- 2026AdaGaR: Adaptive Gabor Representation for Dynamic Scene Reconstructioncited by 8
- 2026World Reconstruction From Inconsistent Viewscited by 8
- 2025Diff4Splat: Controllable 4D Scene Generation with Latent Dynamic Reconstruction Modelscited by 7
- 2026Pantheon360: Taming Digital Twin Generation via 3D-Aware 360deg Video Diffusioncited by 4
- 2026DreamLoop: Controllable Cinemagraph Generation from a Single Photographcited by 2
- 2026AniGS: Bridging Rendering and Diffusion Prior for 3D Scene Animationcited by 1
- 2026Loopy: Seamless Video Loop Generation via Anchored Looping Shift of Positional Embeddingcited by 1
Built on by
Not cited yet.