选择性地提取视觉属性并注入文本到图像模型
Selectively Extracting and Injecting Visual Attributes into Text-to-Image Models
通过优化单个文本词元并引入蒸馏/残差双嵌入，把参考图中的形状、材质、姿态、镜头角度等属性解耦后注入 T2I 生成，可作为无需微调模型即可实现属性级可控生成的即插即用手段。
原文依据 optimizes a text token to selectively represent the target attribute using a custom training prompt and two novel embeddings
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图像模型正日益被用于设计工作流，但仅通过文本表达细微的设计意图仍然是一个挑战。本工作提出一种方法，从参考图像中提取视觉属性并将其直接注入生成流程。具体而言，该方法使用定制的训练提示词和两种新颖的嵌入（称为蒸馏嵌入与残差嵌入）来优化一个文本词元，使其选择性地表示目标属性。通过这一途径可以提取范围广泛的属性，包括形状、材质、姿态以及镜头景别与角度。该方法在取自一个新构建数据集的多种目标属性与文本提示词上得到了有效性验证。在跨多样上下文选择性地提取并应用目标属性方面，该方法优于现有方法。最终，所提方法实现了直观且可控的文本到图像生成，从而简化了设计流程。
英文原文摘要 981 字符
Text-to-image models are increasingly utilized in design workflows, but articulating nuanced design intentions solely through text remains a challenge. This work proposes a method that extracts visual attributes from a reference image and injects them directly into the generation pipeline. Specifically, the method optimizes a text token to selectively represent the target attribute using a custom training prompt and two novel embeddings, termed distilled and residual. A wide range of attributes can be extracted through this approach, including shape, material, pose, and camera shot and angle. The effectiveness of the method is validated on various target attributes and text prompts drawn from a newly constructed dataset. The method outperforms existing approaches in selectively extracting and applying target attributes across diverse contexts. Ultimately, the proposed method enables intuitive and controllable text-to-image generation, streamlining the design process.