Research map / Video / Video generation
LivingWorld: Interactive 4D World Generation with Environmental Dynamics
Hyeongju Mun, In-Hwan Jin, Sohyeong Kim, Kyeongbo Kong
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-aware alignment instead of naive scene flow accumulation (ablation) | LivingWorld · Alignment ablation (WonderWorld) | Global cosine similarity: 0.54 -> 0.91 | Improved |
| add geometry-aware alignment instead of naive scene flow accumulation (ablation) | LivingWorld · Alignment ablation (WonderWorld) | Magnitude ratio: 0.72 -> 0.84 | Improved |
| add geometry-aware alignment instead of naive scene flow accumulation (ablation) | LivingWorld · Alignment ablation (WonderWorld) | MCA: 0.0550 -> 0.0742 | Improved |
| add geometry-aware alignment instead of naive scene flow accumulation (ablation) | LivingWorld · Alignment ablation (WonderWorld) | FMV (lower better): 1.91 -> 0.29 | Improved |
| use geometry-aware alignment instead of 3D-MOM optimization (ablation) | LivingWorld · Alignment ablation (WonderWorld) | Alignment runtime (s): 600.0 -> 12.1 | Improved |
| use geometry-aware alignment instead of 3D-MOM optimization (ablation) | LivingWorld · Alignment ablation (WonderWorld) | Global cosine similarity: 0.66 -> 0.91 | Improved |
| use geometry-aware alignment instead of 3D-MOM optimization (ablation) | LivingWorld · Alignment ablation (WonderWorld) | FMV (lower better): 1.66 -> 0.29 | Improved |
| add bidirectional motion propagation instead of forward-only (ablation) | LivingWorld · Boundary hole analysis | Hole rate (%): 9.8 -> 0.9 | Improved |
| add learned hash-based motion field instead of applying scene flow directly to Gaussians (ablation) | LivingWorld · Qualitative motion field ablation | Motion realism | Improved |
| use linear opacity schedule instead of exponential or sigmoid schedules (ablation) | LivingWorld · Looped rendering (qualitative) | Temporal continuity | Improved |
| compare LivingWorld to Veo 3.1 | LivingWorld · Environmental dynamics benchmark | Runtime (s): 140 -> 12 | Improved |
| compare LivingWorld to Veo 3.1 | LivingWorld · Environmental dynamics benchmark | Imaging quality: 0.694 -> 0.673 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- PerpetualWonder — PerpetualWonder: Long-horizon Action-conditioned 4D Scene Generation (CVPR 2026) · better on 6 of 6 shared results
- Tora — Tora: Trajectory-oriented Diffusion Transformer for Video Generation (CVPR 2025) · better on 6 of 6 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
- Beyond Pixel Histories: World Models with Persistent 3D State (ICML 2026 · similarity 0.89)
- Persistent Nature: A Generative Model of Unbounded 3D Worlds (CVPR 2023 · similarity 0.89)
- WorldReel: 4D Video Generation with Consistent Geometry and Motion Modeling (CVPR 2026 · similarity 0.88)
- World from Motion: Generative Dynamic Gaussian Reconstruction from Monocular Video (NeurIPS 2026 · similarity 0.87)
- 4DWorldBench: A Comprehensive Evaluation Framework for 3D/4D World Generation Models (CVPR 2026 · similarity 0.87)
- DynamicVerse: A Physically-Aware Multimodal Framework for 4D World Modeling (NeurIPS 2025 · similarity 0.86)
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.
