arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2407.06182cs.CV
收藏

作为流均衡的组合视频生成

Compositional Video Generation as Flow Equalization

National University of Singapore(新加坡国立大学)

Xingyi Yang, Xinchao Wang

AI总结:

针对文本到视频扩散模型中概念影响失衡的问题,本文提出 Vico 框架,通过时空注意力图与最大流估计引导潜变量更新,显著提升组合视频生成与编辑的准确性。

摘要:

大规模文本到视频(Text-to-Video,T2V)扩散模型近期展现出前所未有的能力,能够将自然语言描述转化为令人惊叹且具有照片真实感的视频。尽管成果令人鼓舞,但一个重大挑战仍然存在:这些模型难以充分理解多个概念与动作之间复杂的组合交互。当某些词语对最终视频产生主导性影响、遮蔽其他概念时,这一问题便会出现。为解决该问题,我们提出 Vico,这是一个用于组合视频生成的通用框架,可显式确保所有概念都得到恰当表征。Vico 的核心是分析输入 token 如何影响生成视频,并调整模型以防止任何单一概念占据主导。具体而言,Vico 从所有层提取注意力权重以构建时空注意力图,随后将影响估计为从源文本 token 到视频目标 token 的最大流(max-flow)。尽管在扩散模型中直接计算注意力流通常不可行,我们仍基于子图流设计了一种高效近似,并采用快速向量化实现,从而使流计算可控且可微。通过更新带噪潜变量以均衡这些流,Vico 能够捕捉复杂交互,并因此生成紧密贴合文本描述的视频。我们将该方法应用于多个基于扩散的视频模型,以完成组合式 T2V 和视频编辑。实验结果表明,我们的框架显著提升了生成视频的组合丰富性与准确性。请访问我们的网站:https://adamdad.github.io/vico/。

英文摘要:

Large-scale Text-to-Video (T2V) diffusion models have recently demonstrated unprecedented capability to transform natural language descriptions into stunning and photorealistic videos. Despite the promising results, a significant challenge remains: these models struggle to fully grasp complex compositional interactions between multiple concepts and actions. This issue arises when some words dominantly influence the final video, overshadowing other concepts.To tackle this problem, we introduce \textbf{Vico}, a generic framework for compositional video generation that explicitly ensures all concepts are represented properly. At its core, Vico analyzes how input tokens influence the generated video, and adjusts the model to prevent any single concept from dominating. Specifically, Vico extracts attention weights from all layers to build a spatial-temporal attention graph, and then estimates the influence as the \emph{max-flow} from the source text token to the video target token. Although the direct computation of attention flow in diffusion models is typically infeasible, we devise an efficient approximation based on subgraph flows and employ a fast and vectorized implementation, which in turn makes the flow computation manageable and differentiable. By updating the noisy latent to balance these flows, Vico captures complex interactions and consequently produces videos that closely adhere to textual descriptions. We apply our method to multiple diffusion-based video models for compositional T2V and video editing. Empirical results demonstrate that our framework significantly enhances the compositional richness and accuracy of the generated videos. Visit our website at~\href{https://adamdad.github.io/vico/}{\url{https://adamdad.github.io/vico/}}.

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。

↑