Research map / Video / Video generation
DreamWorld: Geometry-Grounded Video Diffusion for 3D-Consistent World Modeling
Haibo Yang, Yang Chen, Yingwei Pan, Zhineng Chen, Ting Yao, Tao Mei
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add geometry diffusion (generated complete geometry) instead of raw incomplete 3D features (ablation) | DreamWorld · RealEstate10K easy | PSNR: 20.05 -> 23.15 | Improved |
| decouple geometry and appearance instead of joint prediction (ablation) | DreamWorld · RealEstate10K easy | PSNR: 21.87 -> 23.15 | Improved |
| decouple geometry and appearance instead of joint prediction (ablation) | DreamWorld · RealEstate10K easy | LPIPS: 0.256 -> 0.249 | Improved |
| condition on 3D foundation-model features instead of VAE-encoded warped frames (ablation) | DreamWorld baseline · RealEstate10K easy | PSNR: 19.47 -> 20.05 | Improved |
| use explicit 3D features with joint geometry-appearance prediction over VAE baseline (ablation) | DreamWorld baseline · RealEstate10K easy | PSNR: 19.47 -> 21.87 | Improved |
| full DreamWorld vs best competitor Gen3C | DreamWorld · RealEstate10K hard | PSNR: 18.04 -> 20.94 | Improved |
| full DreamWorld vs best competitor Gen3C | DreamWorld · Tanks-and-Temples hard | PSNR: 17.95 -> 19.59 | Improved |
| DreamWorld vs WonderWorld on photo consistency | DreamWorld · WorldScore | Photo Consist.: 67.86 -> 92.70 | Improved |
| DreamWorld vs WonderWorld on camera control | DreamWorld · WorldScore | Cam. Ctrl.: 92.32 -> 81.97 | Hurt |
| DreamWorld vs FantasyWorld joint design on average score | DreamWorld · WorldScore | Avg: 64.43 -> 75.04 | Improved |
Later papers that compared with this method
1 later accepted paper reports DreamWorld-1.3B as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- LaMo: Self-Supervised Latent Motion Priors for Physical Realism in Video Generation (NeurIPS 2026 · better on 2 of 2 shared results)
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Geometry Forcing: Marrying Video Diffusion and 3D Representation for Consistent World Modeling (ICLR 2026 · similarity 0.93)
- DreamCraft3D: Hierarchical 3D Generation with Bootstrapped Diffusion Prior (ICLR 2024 · similarity 0.88)
- PanoWorld: Geometry-Consistent Panoramic Video World Modeling (NeurIPS 2026 · similarity 0.88)
- Diffusion4D: Fast Spatial-temporal Consistent 4D generation via Video Diffusion Models (NeurIPS 2024 · similarity 0.88)
- GeoVideo: Introducing Geometric Regularization into Video Generation Model (NeurIPS 2025 · similarity 0.88)
- ViVid-1-to-3: Novel View Synthesis with Video Diffusion Models (CVPR 2024 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
