研究人员开发了一个名为DCR（Default Completion Repulsion）的新框架，以解决扩散模型中一个常见的问题，即它们难以生成稀有但合理的组合。该方法在生成过程中识别并抑制了对更频繁语义配置的偏见。DCR无需重新训练模型即可运行，提高了代表性不足提示的组合保真度和视觉质量。 AI
影响 这项研究为改进扩散模型生成稀有组合的可控性提供了一种新颖的方法。
排序理由 这是一篇详细介绍扩散模型新方法的学术论文。
AI 生成摘要 · Google Gemini · 来自 2 个来源。 我们如何撰写摘要 →
arXiv:2605.06512v1 Announce Type: new Abstract: Diffusion models generate realistic visual content, yet often fail to produce rare but plausible compositions. When prompted with combinations that are valid but underrepresented in training data, such as a snowy beach or a rainbow …
Diffusion models generate realistic visual content, yet often fail to produce rare but plausible compositions. When prompted with combinations that are valid but underrepresented in training data, such as a snowy beach or a rainbow at night, the generation process frequently coll…