English(EN) Multimodal 3D LUT Generation via StatLUT with Statistical Features for Photorealistic Style Transfer

StatLUT 框架支持文本驱动的照片级真实感风格迁移

作者 PulseAugur 编辑部 · [2 个来源] · 2026-07-09 08:23

研究人员推出了一种新颖的多模态框架 StatLUT，用于生成用于照片级真实感风格迁移的 3D 查找表 (LUT)。该框架通过使用 Lab-Extractor 将颜色风格与结构语义分离，解决了现有深度学习方法的局限性，从而防止了不自然的失真。StatLUT 将 LUT 生成构建为基于 Transformer 的 Seq2Seq 翻译任务，并使用多维残差映射器来确保拓扑平滑的 LUT。此外，它还集成了 H-Diffuser，一个扩散 Transformer，通过合成自然语言提示的特征来实现文本驱动的色彩分级。 AI

影响这项研究可能带来更直观、无伪影的图像风格迁移，对创意工具和媒体制作产生潜在影响。

排序理由该集群包含一篇详细介绍图像风格迁移新框架的研究论文。

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 →

报道来源 [2]

arXiv cs.CV TIER_1 English(EN) · Yifan Wang, Zhixiang Hao, Yu Wang, Congchao Zhu · 2026-07-10 04:00

Multimodal 3D LUT Generation via StatLUT with Statistical Features for Photorealistic Style Transfer

arXiv:2607.08227v1 Announce Type: new Abstract: Photorealistic Style Transfer (PST) aims to transfer the color and tonal style of a reference to a content image while strictly preserving its structural integrity. However, existing deep learning-based methods inherently suffer fro…
arXiv cs.CV TIER_1 English(EN) · Congchao Zhu · 2026-07-09 08:23

Multimodal 3D LUT Generation via StatLUT with Statistical Features for Photorealistic Style Transfer

Photorealistic Style Transfer (PST) aims to transfer the color and tonal style of a reference to a content image while strictly preserving its structural integrity. However, existing deep learning-based methods inherently suffer from semantic entanglement caused by pre-trained im…

报道来源 [2]

Multimodal 3D LUT Generation via StatLUT with Statistical Features for Photorealistic Style Transfer

Multimodal 3D LUT Generation via StatLUT with Statistical Features for Photorealistic Style Transfer

相关实体

相关话题