无需训练即可融合多个主体的LoRA模块，实现多主体个性化图像生成。
FreeLoRA: Enabling Training-Free LoRA Fusion for Autoregressive Multi-Subject Personalization
- 采用全标记调优策略，仅用少量图像适配每个主体的LoRA模块。
- 推理时仅激活对应主体的模块，实现多主体共存且不互相干扰。
- 适合需要快速部署多主体个性化生成的场景，如虚拟试衣、海报设计。
主体驱动的图像生成在虚拟试穿和海报设计等应用中至关重要。现有方法通常对预训练生成模型进行微调或使用LoRA进行个体主体适配，但难以实现多主体个性化，因独立适配模块的组合常需复杂重调优或联合优化。我们提出FreeLoRA，一种简单且通用的框架，实现训练-free的主体专属LoRA模块融合。每个LoRA模块通过少量特定主体图像，采用全标记调优（Full Token Tuning）策略进行适配，该策略将模块应用于提示中的所有标记，以促进弱监督下的标记-内容对齐。推理时，采用主体感知推理（Subject-Aware Inference），仅在对应主体标记上激活相应模块。这使得单张图像中可无训练融合多个个性化主体，同时缓解过拟合与主体间相互干扰问题。大量实验表明，FreeLoRA在主体保真度和提示一致性方面均表现优异。
原文摘要 · Abstract (English)
Subject-driven image generation plays a crucial role in applications such as virtual try-on and poster design. Existing approaches typically fine-tune pretrained generative models or apply LoRA-based adaptations for individual subjects. However, these methods struggle with multi-subject personalization, as combining independently adapted modules often requires complex re-tuning or joint optimization. We present FreeLoRA, a simple and generalizable framework that enables training-free fusion of subject-specific LoRA modules for multi-subject personalization. Each LoRA module is adapted on a few images of a specific subject using a Full Token Tuning strategy, where it is applied across all tokens in the prompt to encourage weakly supervised token-content alignment. At inference, we adopt Subject-Aware Inference, activating each module only on its corresponding subject tokens. This enables training-free fusion of multiple personalized subjects within a single image, while mitigating overfitting and mutual interference between subjects. Extensive experiments show that FreeLoRA achieves strong performance in both subject fidelity and prompt consistency.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。