PulseAugur
实时 06:15:46
English(EN) Don't Settle at the Mode! Mitigating Diversity Collapse in Pretrained Flow Models via Feature Self-Guidance

新的自引导方法提高了AI图像生成的图像多样性

研究人员开发了一种新的无需训练的方法,称为特征自引导,以解决用于图像生成的预训练流模型中的多样性崩溃问题。该技术在批量生成期间分散内部特征,并使用流形正则化使它们与数据流形对齐,从而在不牺牲质量的情况下确保输出的多样性。这种即插即用的模块仅带来边际推理成本,并在文本到图像和深度到图像生成等各种条件流模型的图像多样性方面显示出显著的改进。 AI

影响 增强了AI生成图像的多样性和质量,可能改进创意领域和内容生成中的应用。

排序理由 该集群包含一篇详细介绍改进AI模型性能新方法的学术论文。

在 arXiv cs.CV 阅读 →

AI 生成摘要 · Google Gemini · 来自 2 个来源。 我们如何撰写摘要 →

新的自引导方法提高了AI图像生成的图像多样性

报道来源 [2]

  1. arXiv cs.CV TIER_1 English(EN) · Pradhaan S Bhat, Rishubh Parihar, Abhijnya Bhat, R. Venkatesh Babu ·

    Don't Settle at the Mode! Mitigating Diversity Collapse in Pretrained Flow Models via Feature Self-Guidance

    arXiv:2606.27371v1 Announce Type: new Abstract: State-of-the-art flow models generate stunning images from text or image prompts. However, they suffer from diversity collapse when generating multiple samples under the same conditioning. Existing methods address this issue via eit…

  2. arXiv cs.CV TIER_1 English(EN) · R. Venkatesh Babu ·

    Don't Settle at the Mode! Mitigating Diversity Collapse in Pretrained Flow Models via Feature Self-Guidance

    State-of-the-art flow models generate stunning images from text or image prompts. However, they suffer from diversity collapse when generating multiple samples under the same conditioning. Existing methods address this issue via either latent guidance, which has limited effective…