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 2603.29569cs.GR
收藏

AdaptDiff: 差分模型中适应性引导以生成多样且身份一致的面部合成

AdaptDiff: Adaptive Guidance in Diffusion Models for Diverse and Identity-Consistent Face Synthesis (Student Abstract)

Fraunhofer IGD(弗劳恩霍夫应用研究促进协会IGD);Department of Computer Science, TU Darmstadt(达姆施塔特工业大学计算机系)

Eduarda Caldeira, Tahar Chettaoui, Naser Damer, Fadi Boutros

AI总结:

本文提出动态加权方案,通过适应性引导抑制无关属性,提升生成面部的一致性和多样性。

摘要:

差分模型基于身份嵌入进行条件生成,可生成在多个样本中一致保持身份的合成面部图像。近期研究表明,通过分类器自由引导在采样过程中引入额外的负条件,可提供抑制不期望属性的机制,从而提高类间分离度。基于此见解,我们提出一种动态加权方案,用于负条件在整个采样轨迹中适应。该策略利用正负条件在生成不同阶段的互补优势,导致更加多样但身份一致的合成数据。

英文摘要:

Diffusion models conditioned on identity embeddings enable the generation of synthetic face images that consistently preserve identity across multiple samples. Recent work has shown that introducing an additional negative condition through classifier-free guidance during sampling provides a mechanism to suppress undesired attributes, thus improving inter-class separability. Building on this insight, we propose a dynamic weighting scheme for the negative condition that adapts throughout the sampling trajectory. This strategy leverages the complementary strengths of positive and negative conditions at different stages of generation, leading to more diverse yet identity-consistent synthetic data.

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。

↑