研究人员开发了WithEveryone，一个旨在改进包含多个角色的多人图像生成的新颖框架。该系统解决了当前模型在保持个体身份及其在场景中位置方面的不可靠性问题，尤其是在生成多达十个人的图像时。通过将身份锚定到布局规划并采用基于区域的身份损失，WithEveryone增强了身份相似性，并与GPT-Image-2等现有方法相比显著减少了伪影。 AI
影响 该框架可能带来更可靠、更高质量的AI生成多人图像，影响创意工具和应用。
排序理由 该集群描述了一篇关于用于图像生成的新颖框架的最新研究论文。
在 Hugging Face Daily Papers 阅读 →
AI 生成摘要 · Google Gemini · 来自 3 个来源。 我们如何撰写摘要 →
Identity-preserving image generation becomes increasingly unreliable when a scene must contain many specified people. Beyond retaining each identity, the model must bind every reference to a distinct person and location, while training-time identity losses must establish correspo…
WithEveryone enables reliable identity-preserving group image generation for up to ten people by grounding identities to layout plans and using region-based identity losses.
arXiv:2608.20336v1 Announce Type: new Abstract: Identity-preserving image generation becomes increasingly unreliable when a scene must contain many specified people. Beyond retaining each identity, the model must bind every reference to a distinct person and location, while train…