SwitchCraft：基于注意力控制的免训练多事件视频生成
SwitchCraft: Training-Free Multi-Event Video Generation with Attention Controls
EAQS 通过按事件对齐帧级注意力来引导提示词注入，ABSS 自适应调节引导强度，这套免训练的注意力操控思路可直接迁移到任何 DiT 类视频扩散模型上，用于解决多阶段叙事、镜头切换等时序可控生成问题。
原文依据 we introduce Event-Aligned Query Steering (EAQS), which steers frame-level attention to align with relevant event prompts
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论文本到视频扩散模型的最新进展已实现高保真且时序连贯的视频合成。然而，当前模型主要针对单事件生成进行优化。在处理多事件提示词时，由于缺乏显式的时序定位，此类模型往往产生混合或坍缩的场景，破坏了预期的叙事结构。为解决这一局限，我们提出 SwitchCraft，一个免训练的多事件视频生成框架。我们的关键洞见是：在时间上均匀注入提示词忽略了事件与帧之间的对应关系。为此，我们引入事件对齐查询引导（EAQS），将帧级注意力引导至与相关事件提示词对齐。此外，我们提出自动平衡强度求解器（ABSS），自适应地平衡引导强度，以保持时序一致性与视觉保真度。大量实验表明，与现有基线相比，SwitchCraft 显著提升了提示词对齐、事件清晰度与场景一致性，为多事件视频生成提供了一种简单而有效的解决方案。
英文原文摘要 1119 字符
Recent advances in text-to-video diffusion models have enabled high-fidelity and temporally coherent video synthesis. However, current models are predominantly optimized for single-event generation. When handling multi-event prompts, without explicit temporal grounding, such models often produce blended or collapsed scenes that break the intended narrative. To address this limitation, we present SwitchCraft, a training-free framework for multi-event video generation. Our key insight is that uniform prompt injection across time ignores the correspondence between events and frames. To this end, we introduce Event-Aligned Query Steering (EAQS), which steers frame-level attention to align with relevant event prompts. Furthermore, we propose Auto-Balance Strength Solver (ABSS), which adaptively balances steering strength to preserve temporal consistency and visual fidelity. Extensive experiments demonstrate that SwitchCraft substantially improves prompt alignment, event clarity, and scene consistency compared with existing baselines, offering a simple yet effective solution for multi-event video generation.