两篇新研究论文探讨了在无需模型重新训练的情况下控制 AI 生成图像颜色的方法。第一篇“Colorful-Noise”操纵扩散模型初始噪声的低频分量，以影响全局结构和颜色。第二篇“Color Conditional Generation with Sliced Wasserstein Guidance”采用一种无需训练的方法，根据参考图像的颜色分布来引导扩散过程，旨在保持语义一致性。 AI
影响 引入了新的无需训练的技术，以增强扩散模型中的颜色控制，可能提高图像生成的真实感和用户定制性。
排序理由 两篇学术论文发表在 arXiv 上，提出了用于图像生成颜色控制的新颖方法。
AI 生成摘要 · Google Gemini · 来自 4 个来源。 我们如何撰写摘要 →
arXiv:2512.21788v3 Announce Type: replace Abstract: Parameter-Efficient Fine-Tuning of Diffusion Transformers (DiTs) for diverse, multi-conditional tasks often suffers from task interference when using monolithic adapters like LoRA. The Mixture of Low-rank Experts (MoLE) architec…
arXiv:2605.00548v1 Announce Type: new Abstract: Text-to-image diffusion models generate images by gradually converting white Gaussian noise into a natural image. White Gaussian noise is well suited for producing diverse outputs from a single text prompt due to its absence of stru…
arXiv:2503.19034v2 Announce Type: replace Abstract: We propose SW-Guidance, a training-free approach for image generation conditioned on the color distribution of a reference image. While it is possible to generate an image with fixed colors by first creating an image from a text…
Text-to-image diffusion models generate images by gradually converting white Gaussian noise into a natural image. White Gaussian noise is well suited for producing diverse outputs from a single text prompt due to its absence of structure. However, this very property limits contro…