DCoAR: 一种深度概念注入到统一自回归模型中用于个性化文本到图像生成
DCoAR: Deep Concept Injection into Unified Autoregressive Models for Personalized Text-to-Image Generation
AI总结：
DCoAR通过深度概念注入框架实现个性化文本到图像生成，采用LMCL策略和多正则化方案提升生成质量与上下文适应性。
摘要：
统一自回归（AR）模型在多模态理解和生成方面表现出色。然而，其在定制图像生成领域的全部潜力尚未完全实现。现有的统一AR模型定制方法面临一个根本性困境：基于适应的方法容易过拟合且存在可扩展性瓶颈，而基于概念注入的范式则受限于浅层注入策略，导致视觉保真度差且上下文重新情境化能力受损。为此，我们提出了DCoAR，一种新的深度概念注入框架，该框架保持完全冻结的预训练模型。DCoAR通过分层多模态上下文学习（LMCL）策略深度整合新概念，该策略通过多方面的正则化方案稳定：双重先验保持（DPP）损失用于缓解语义漂移，以及上下文感知自正则化（CASR）损失用于增强上下文重新情境化能力。该框架还支持在用户提供的风格中无训练主体定制。实验表明，DCoAR显著优于先前的注入方法，并在可训练参数大幅减少的情况下实现了与基于适应方法相当的性能。代码：https://github.com/KZF-kzf/CoAR
英文摘要：
The unified autoregressive (AR) model excels at multimodal understanding and generation. However, its full potential in the domain of customized image generation has yet to be fully realized. Existing customization approaches for unified AR models face a fundamental dilemma: adaptation-based methods suffer from overfitting and scalability bottlenecks, while concept-injection paradigms are constrained by a shallow injection strategy that leads to poor visual fidelity and impaired re-contextualization. To address this, we propose DCoAR, a novel deep concept injection framework that maintains a completely frozen pre-trained model. DCoAR deeply integrates new concepts through a Layer-wise Multimodal Context Learning (LMCL) strategy, which is stabilized by a multi-faceted regularization scheme: a Dual Prior Preservation (DPP) loss to mitigate semantic drift and a Context-Aware Self-Regularization (CASR) loss to enhance re-contextualization. The framework also enables training-free subject customization in user-provided styles. Experiments demonstrate that DCoAR significantly outperforms previous injection-based methods and achieves performance competitive with adaptation-based approaches while requiring substantially fewer trainable parameters. Code: https://github.com/KZF-kzf/CoAR
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。
