Research map / Video / Video generation
World-R1: Reinforcing 3D Constraints for Text-to-Video Generation
Weijie Wang, Xiaoxuan He, Youping Gu, Yifan Yang, Zeyu Zhang, Yefei He, Yanbo Ding, Xirui Hu, Donny Y. Chen, Zhiyuan He, Yuqing Yang, Bohan Zhuang
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 |
|---|---|---|---|
| fine-tune with World-R1 RL alignment (Small) | Wan2.1-T2V-1.3B · 3D consistency | PSNR: 17.40 -> 27.63 | Improved |
| fine-tune with World-R1 RL alignment (Large) | Wan2.1-T2V-14B · 3D consistency | PSNR: 19.76 -> 27.67 | Improved |
| fine-tune with World-R1 RL alignment (Small) | Wan2.1-T2V-1.3B · 3D consistency | LPIPS: 0.467 -> 0.201 | Improved |
| fine-tune with World-R1 RL alignment (Small) | Wan2.1-T2V-1.3B · VBench | Aesthetic Quality: 62.43 -> 65.74 | Improved |
| fine-tune with World-R1 RL alignment (Small) | Wan2.1-T2V-1.3B · VBench | Background Consistency: 97.29 -> 96.67 | Hurt |
| fine-tune with World-R1 RL alignment (Small) | Wan2.1-T2V-1.3B · VBench | Motion Smoothness: 97.44 -> 98.55 | Improved |
| add the 3D-aware reward R_3D (ablation) | World-R1-Small · reward ablation | geometric consistency | Improved |
| add the general generation reward R_gen (ablation) | World-R1-Small · reward ablation | perceptual fidelity | Improved |
| add implicit camera injection via noise warping (ablation) | World-R1-Small · convergence and trajectory alignment | convergence speed and trajectory alignment | Improved |
| add periodic decoupled dynamic fine-tuning (ablation) | World-R1-Small · dynamic motion | non-rigid dynamics | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- ReCamMaster — ReCamMaster: Camera-Controlled Generative Rendering from A Single Video (ICCV 2025) · better on 4 of 5 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
- WorldForge: Forging Unified World Modeling into Video Generation (NeurIPS 2026 · similarity 0.89)
- Let Language Constrain Geometry: Vision–Language Models as Semantic and Spatial Critics for 3D Generation (ICML 2026 · similarity 0.88)
- SimVS: Simulating World Inconsistencies for Robust View Synthesis (CVPR 2025 · similarity 0.87)
- WorldWeaver: Generating Long-Horizon Video Worlds via Rich Perception (NeurIPS 2025 · similarity 0.87)
- Video World Models with Long-term Spatial Memory (NeurIPS 2025 · similarity 0.87)
- 4DWorldBench: A Comprehensive Evaluation Framework for 3D/4D World Generation Models (CVPR 2026 · 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.
