Di3PO - Diptych Diffusion DPO用于图像生成的针对性改进
Di3PO - Diptych Diffusion DPO for Targeted Improvements in Image Generation
AI总结：
Di3PO通过针对性改进图像生成中的特定区域，提升扩散模型的文本渲染性能。
摘要：
现有方法在文本到图像（T2I）扩散模型的偏好调优中往往依赖计算成本高昂的生成步骤来创建正负图像对。这些方法经常产生缺乏有意义差异、采样和过滤成本高或在无关像素区域存在显著变化的训练对，从而降低训练效率。为了解决这些限制，我们引入了
英文摘要：
Existing methods for preference tuning of text-to-image (T2I) diffusion models often rely on computationally expensive generation steps to create positive and negative pairs of images. These approaches frequently yield training pairs that either lack meaningful differences, are expensive to sample and filter, or exhibit significant variance in irrelevant pixel regions, thereby degrading training efficiency. To address these limitations, we introduce "Di3PO", a novel method for constructing positive and negative pairs that isolates specific regions targeted for improvement during preference tuning, while keeping the surrounding context in the image stable. We demonstrate the efficacy of our approach by applying it to the challenging task of text rendering in diffusion models, showcasing improvements over baseline methods of SFT and DPO.
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。
