arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2412.04460cs.CV
收藏

LayerFusion:基于生成先验的协调多层文本到图像生成

LayerFusion: Harmonized Multi-Layer Text-to-Image Generation with Generative Priors

Virginia Tech(弗吉尼亚理工大学);Adobe Research(奥多比研究院)

Yusuf Dalva, Yijun Li, Qing Liu, Nanxuan Zhao, Jianming Zhang, Zhe Lin, Pinar Yanardag

AI总结:

针对分层内容生成未充分探索的问题,提出基于潜在扩散模型的LayerFusion管道,通过协调生成机制实现前景层(RGBA)与背景层(RGB)的动态交互,生成结果在视觉连贯性等方面显著优于基线方法。

摘要:

大规模扩散模型在从文本描述生成高质量图像方面取得了显著成功,在各类应用中广受欢迎。然而,分层内容(如包含前景和背景层的透明图像)的生成仍是一个未充分探索的领域。分层内容生成对于平面设计、动画和数字艺术等领域的创意工作流至关重要,这些领域中基于层的方法是灵活编辑和合成的基础。本文提出一种基于潜在扩散模型(LDMs)的新型图像生成管道,可生成包含两层的图像:带有透明度信息的前景层(RGBA)和背景层(RGB)。与现有顺序生成这些层的方法不同,我们的方法引入了协调生成机制,使层间能够动态交互以获得更连贯的输出。通过大量定性和定量实验,我们证明了该方法的有效性,与基线方法相比,在视觉连贯性、图像质量和层一致性方面均有显著提升。

英文摘要:

Large-scale diffusion models have achieved remarkable success in generating high-quality images from textual descriptions, gaining popularity across various applications. However, the generation of layered content, such as transparent images with foreground and background layers, remains an under-explored area. Layered content generation is crucial for creative workflows in fields like graphic design, animation, and digital art, where layer-based approaches are fundamental for flexible editing and composition. In this paper, we propose a novel image generation pipeline based on Latent Diffusion Models (LDMs) that generates images with two layers: a foreground layer (RGBA) with transparency information and a background layer (RGB). Unlike existing methods that generate these layers sequentially, our approach introduces a harmonized generation mechanism that enables dynamic interactions between the layers for more coherent outputs. We demonstrate the effectiveness of our method through extensive qualitative and quantitative experiments, showing significant improvements in visual coherence, image quality, and layer consistency compared to baseline methods.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑