WorldReel: 4D Video Generation with Consistent Geometry and Motion Modeling
Shaoheng Fang, Hanwen Jiang, Yunpeng Bai, Niloy J. Mitra, Qixing Huang
摘要
Recent video generators achieve striking photorealism, yet remain fundamentally inconsistent in 3D. We present WorldReel, a 4D video generator that is natively spatio-temporally consistent. WorldReel jointly produces RGB frames together with 4D scene representations, including pointmaps, camera trajectory, and dense flow mapping, enabling coherent geometry and appearance modeling over time. Our explicit 4D representation enforces a single underlying scene that persists across viewpoints and dynamic content, yielding videos that remain consistent even under large non-rigid motion and significant camera movement.We train WorldReel by carefully combining synthetic and real data: synthetic data provides precise 4D supervision (geometry, motion, and camera), while real videos contribute visual diversity and realism. This blend allows WorldReel to generalize to in-the-wild footage while preserving strong geometric fidelity.Extensive experiments demonstrate that WorldReel sets a new state-of-the-art for consistent video generation with dynamic scenes and moving cameras, improving metrics of geometric consistency, motion coherence, and reducing view-time artifacts over competing methods. We believe that WorldReel brings video generation closer to 4D-consistent world modeling, where agents can render, interact, and reason about scenes through a single and stable spatiotemporal representation.
问问这篇 Paper
智能体会读完全文。
Lune 把这篇 Paper 索引到了每一个公式,引用它的顶会 Paper 也一样。你提问,回答直接引用原文。
它引用的顶会 Paper51
- High-Resolution Image Synthesis with Latent Diffusion ModelsRobin Rombach, Andreas Blattmann, Dominik Lorenz, Patrick Esser 等CVPR 2022 · 被引用 13,123 次
- Photorealistic Text-to-Image Diffusion Models with Deep Language UnderstandingChitwan Saharia, William Chan, Saurabh Saxena, Lala Li 等NeurIPS 2022 · 被引用 8,965 次
- Scalable Diffusion Models with TransformersWilliam Peebles, Saining XieICCV 2023 · 被引用 5,568 次
- Video Diffusion ModelsJonathan Ho, Tim Salimans, Alexey A. Gritsenko, William Chan 等NeurIPS 2022 · 被引用 2,948 次
- Vision Transformers for Dense PredictionRené Ranftl, Alexey Bochkovskiy, Vladlen KoltunICCV 2021 · 被引用 2,647 次
相关 Paper
- Geometry-aware 4D Video Generation for Robot ManipulationZeyi Liu, Shuang Li, Eric Cousineau, Siyuan Feng 等ICLR 2026 · 被引用 28 次
- Restage4D: Reanimating Deformable 3D Reconstruction from a Single VideoJixuan He, Chieh Hubert Lin, Lu Qi, Ming-Hsuan YangNeurIPS 2025 · 被引用 2 次
- ShapeGen4D: Towards High Quality 4D Shape Generation from VideosJiraphon Yenphraphai, Ashkan Mirzaei, Jianqi Chen, Jiaxu Zou 等ICLR 2026 · 被引用 21 次
- VerseCrafter: Dynamic Realistic Video World Model with 4D Geometric ControlSixiao Zheng, Minghao Yin, Wenbo Hu, Xiaoyu Li 等CVPR 2026 · 被引用 27 次
- Generating Long Videos of Dynamic ScenesTim Brooks, Janne Hellsten, Miika Aittala, Ting-Chun Wang 等NeurIPS 2022 · 被引用 152 次
