上下文空间中的即时排斥以实现扩散变换器的丰富多样性
On-the-fly Repulsion in the Contextual Space for Rich Diversity in Diffusion Transformers
AI总结：
针对文本到图像扩散模型多样性不足的问题，提出在扩散变换器的上下文空间中通过多模态注意力通道施加即时排斥，在不牺牲视觉保真度和语义一致性的前提下显著提升生成多样性，且计算开销小，适用于现代Turbo和蒸馏模型。
摘要：
现代文本到图像（T2I）扩散模型在语义对齐方面取得了显著进展，但通常缺乏多样性，倾向于为任何给定提示收敛到狭窄的视觉解决方案集。这种典型性偏差对需要广泛生成结果的创意应用构成了挑战。我们识别出当前多样性方法中的一个基本权衡：修改模型输入需要昂贵的优化来整合生成路径的反馈。相反，对空间上已承诺的中间潜变量进行操作往往会破坏正在形成的视觉结构，导致伪影。在这项工作中，我们提出在上下文空间中应用排斥作为一种新颖的框架，以实现扩散变换器的丰富多样性。通过干预多模态注意力通道，我们在变换器的前向传播过程中施加即时排斥，在文本条件被新兴图像结构丰富后的块之间注入干预。这允许在结构信息形成后但构图固定之前重定向引导轨迹。我们的结果表明，上下文空间中的排斥在不牺牲视觉保真度或语义一致性的情况下产生了显著更丰富的多样性。此外，我们的方法非常高效，计算开销小，即使在现代“Turbo”和蒸馏模型中也有效，而传统的基于轨迹的干预在这些模型中通常会失败。
英文摘要：
Modern Text-to-Image (T2I) diffusion models have achieved remarkable semantic alignment, yet they often suffer from a significant lack of variety, converging on a narrow set of visual solutions for any given prompt. This typicality bias presents a challenge for creative applications that require a wide range of generative outcomes. We identify a fundamental trade-off in current approaches to diversity: modifying model inputs requires costly optimization to incorporate feedback from the generative path. In contrast, acting on spatially-committed intermediate latents tends to disrupt the forming visual structure, leading to artifacts. In this work, we propose to apply repulsion in the Contextual Space as a novel framework for achieving rich diversity in Diffusion Transformers. By intervening in the multimodal attention channels, we apply on-the-fly repulsion during the transformer's forward pass, injecting the intervention between blocks where text conditioning is enriched with emergent image structure. This allows for redirecting the guidance trajectory after it is structurally informed but before the composition is fixed. Our results demonstrate that repulsion in the Contextual Space produces significantly richer diversity without sacrificing visual fidelity or semantic adherence. Furthermore, our method is uniquely efficient, imposing a small computational overhead while remaining effective even in modern "Turbo" and distilled models where traditional trajectory-based interventions typically fail.
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。
