Infinity-RoPE：基于自回归自展开的可控无限视频生成技术

Infinity-RoPE: Action-Controllable Infinite Video Generation Emerges From Autoregressive Self-Rollout

November 25, 2025

作者: Hidir Yesiltepe, Tuna Han Salih Meral, Adil Kaan Akan, Kaan Oktay, Pinar Yanardag

cs.AI

摘要

当前自回归视频扩散模型面临三大核心瓶颈：(i)基础模型的三维旋转位置编码(3D-RoPE)所施加的有限时间跨度限制；(ii)生成长序列视频时提示词响应迟缓导致细粒度动作控制能力下降；(iii)无法在单次生成流中实现非连续的电影式场景转换。我们提出infty-RoPE这一统一推理框架，通过三个相互关联的组件——块相对论RoPE、KV刷新和RoPE截断——系统解决上述局限。块相对论RoPE将时间编码重构为移动局部参考系：新生成的潜在块相对于基础模型最大帧范围进行旋转，而早期块则反向旋转以保持相对时间几何关系。这种相对论表述消除了固定时间位置，实现了远超基础位置限制的连续视频生成。为实现无需重新编码的细粒度动作控制，KV刷新机制通过仅保留全局锚点帧与最后生成的潜在帧来更新KV缓存，从而确保即时提示词响应。最后，RoPE截断技术在时间RoPE坐标中引入受控间断，使得单次连续生成过程中可实现多镜头场景转换。这些组件共同构成了infty-RoPE——一个无需训练即可实现无限时长、可控且具电影感的视频扩散基础框架。综合实验表明，infty-RoPE在VBench各项指标上均持续超越现有自回归模型。

English

Current autoregressive video diffusion models are constrained by three core bottlenecks: (i) the finite temporal horizon imposed by the base model's 3D Rotary Positional Embedding (3D-RoPE), (ii) slow prompt responsiveness in maintaining fine-grained action control during long-form rollouts, and (iii) the inability to realize discontinuous cinematic transitions within a single generation stream. We introduce infty-RoPE, a unified inference-time framework that addresses all three limitations through three interconnected components: Block-Relativistic RoPE, KV Flush, and RoPE Cut. Block-Relativistic RoPE reformulates temporal encoding as a moving local reference frame, where each newly generated latent block is rotated relative to the base model's maximum frame horizon while earlier blocks are rotated backward to preserve relative temporal geometry. This relativistic formulation eliminates fixed temporal positions, enabling continuous video generation far beyond the base positional limits. To obtain fine-grained action control without re-encoding, KV Flush renews the KV cache by retaining only two latent frames, the global sink and the last generated latent frame, thereby ensuring immediate prompt responsiveness. Finally, RoPE Cut introduces controlled discontinuities in temporal RoPE coordinates, enabling multi-cut scene transitions within a single continuous rollout. Together, these components establish infty-RoPE as a training-free foundation for infinite-horizon, controllable, and cinematic video diffusion. Comprehensive experiments show that infty-RoPE consistently surpasses previous autoregressive models in overall VBench scores.