统一的个性化理解、生成与编辑
Unified Personalized Understanding, Generating and Editing
AI总结：
OmniPersona是首个统一LMMs的端到端个性化框架，通过结构解耦的概念标记和显式知识重播机制，实现个性化理解、生成和图像编辑的一致可控。
摘要：
统一的大多模态模型（LMMs）在通用多模态理解和生成方面取得了显著进展。然而，它们仍然遵循一种"一刀切"的范式，难以在一致且可控的方式下建模用户特定的概念（例如生成一张\texttt{<maeve>}的照片）。现有的个性化方法通常依赖于外部检索，这效率低下且难以整合到统一的多模态流水线中。最近的个性化统一模型引入了可学习的软提示来编码概念信息，但它们要么将理解和生成耦合在一起，要么依赖于复杂的多阶段训练，导致跨任务干扰，最终导致模糊或不一致的个性化知识。我们提出了\textbf{OmniPersona}，这是一个为统一LMMs设计的端到端个性化框架，首次在单一架构中集成了个性化理解和生成以及图像编辑。OmniPersona引入了结构解耦的概念标记，为不同任务分配专用子空间以最小化干扰，并结合了显式知识重播机制，该机制在任务间传播个性化属性知识，从而实现一致的个性化行为。为了系统评估统一的个性化，我们提出了\textbf{\texttt{OmniPBench}}，扩展了公开的UnifyBench概念集，加入个性化编辑任务和跨任务评估协议，整合理解和生成与编辑。实验结果表明，OmniPersona在多样化的个性化任务中表现出竞争性和稳健性。我们希望OmniPersona能成为强大的基准，并推动进一步的研究，以可控的统一个性化为目标。
英文摘要：
Unified large multimodal models (LMMs) have achieved remarkable progress in general-purpose multimodal understanding and generation. However, they still operate under a ``one-size-fits-all'' paradigm and struggle to model user-specific concepts (e.g., generate a photo of \texttt{<maeve>}) in a consistent and controllable manner. Existing personalization methods typically rely on external retrieval, which is inefficient and poorly integrated into unified multimodal pipelines. Recent personalized unified models introduce learnable soft prompts to encode concept information, yet they either couple understanding and generation or depend on complex multi-stage training, leading to cross-task interference and ultimately to fuzzy or misaligned personalized knowledge. We present \textbf{OmniPersona}, an end-to-end personalization framework for unified LMMs that, for the first time, integrates personalized understanding, generation, and image editing within a single architecture. OmniPersona introduces structurally decoupled concept tokens, allocating dedicated subspaces for different tasks to minimize interference, and incorporates an explicit knowledge replay mechanism that propagates personalized attribute knowledge across tasks, enabling consistent personalized behavior. To systematically evaluate unified personalization, we propose \textbf{\texttt{OmniPBench}}, extending the public UnifyBench concept set with personalized editing tasks and cross-task evaluation protocols integrating understanding, generation, and editing. Experimental results demonstrate that OmniPersona delivers competitive and robust performance across diverse personalization tasks. We hope OmniPersona will serve as a strong baseline and spur further research on controllable, unified personalization.
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。
