MV-Forcing：基于4D锚定时空自强制的长时长多视角视频生成
MV-Forcing: Long Multi-View Video Generation via 4D-Grounded Spatio-Temporal Self-Forcing
AI总结：
针对动态场景长时多视角一致视频生成难题，提出MV-Forcing框架，融合时序与视角自回归，实现任意时长视角数的高一致性多视角视频生成。
摘要：
视频扩散模型的最新进展已可通过时序自回归实现长时长单视角生成，或通过双向注意力实现短时长多视角合成。但动态场景的长时长、多视角一致视频生成问题仍未得到解决。本研究提出MV-Forcing框架，通过在序贯生成的视角之间引入4D几何桥接，在单个扩散模型内组合时序与逐视角自回归机制。核心思路是让自回归3D重建模型在自回归生成的不同视角之间自然衔接：给定已完成的源视角，先重建其3D结构并渲染出下一个目标视点的几何先验，再由扩散模型将该先验优化为高质量视频。为突破教师模型固定时序窗口的生成限制，本研究引入联合去噪机制，训练过程中两个视角槽均从噪声初始化，支持无边界的时序生成。研究通过带时空自强制的分布匹配蒸馏对模型进行蒸馏，消除了时序与视角序贯自回归过程中的训练-推理暴露偏差。在合成数据与真实世界数据上开展的大量实验表明，MV-Forcing仅需单步少步学生模型，即可生成任意时长、任意视角数的几何一致动态场景多视角视频。
英文摘要：
Recent advances in video diffusion models have enabled either long single-view generation through temporal autoregression, or short multi-view synthesis through bidirectional attention. However, generating long, multi-view consistent videos of dynamic scenes remains unsolved. In this work, we present MV-Forcing, a framework that composes temporal and view-wise autoregression within a single diffusion model by introducing a 4D geometric bridge between sequentially generated views. Our key insight is that an autoregressive 3D reconstruction model naturally interfaces between autoregressively generated views. Given a completed source view, we reconstruct its 3D structure and render a geometric prior of the next target viewpoint, which the diffusion model refines into a high-quality video. To extend generation beyond the teacher's fixed temporal window, we introduce a joint denoising regime where both view slots are initialized from noise during training, enabling temporally unbounded generation. We distill the model via Distribution Matching Distillation with Spatio-Temporal Self-Forcing, closing the train-inference exposure bias gap for both temporal and view-sequential autoregression. Extensive experiments on both synthetic and real-world data demonstrate that MV-Forcing produces geometrically consistent multi-view videos of dynamic scenes at arbitrary lengths and viewpoint counts using a single few-step student model.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
