InstantRetouch:基于双边空间的高效高保真指令引导图像润色
InstantRetouch: Efficient and High-Fidelity Instruction-Guided Image Retouching with Bilateral Space
AI总结:
提出一种基于双边空间操作的图像润色方法,通过预测低分辨率双边网格并利用学习引导图切片,结合扩散模型蒸馏和提示对齐损失,实现高效、高保真且遵循指令的图像润色。
摘要:
语言引导的照片润色旨在调整颜色和色调,同时保留几何和纹理。最近,基于扩散的润色显示出优越的视觉质量,但由于其生成性质,常常面临保真度问题,并且由于其迭代采样过程,效率低下。在这项工作中,我们提出了一种高效且保真的润色方法,使用双边空间操作,该方法既紧凑又内容解耦。具体来说,我们的模型不是直接编辑像素或图像潜在表示,而是预测一个低分辨率的仿射变换双边网格,该网格通过学习的引导图进行切片,然后应用于全分辨率图像。这种方法实现了高保真度和更高的效率。为了保留预训练生成模型的强先验,我们使用变分分数蒸馏将多步扩散模型蒸馏到我们的双边网格框架中,并辅以提示对齐损失来指导指令跟随行为。此外,我们引入了一个新的基准,并在多个维度上评估我们的方法:保真度、指令遵循和效率。与最新的润色方法(如Gemini-2.5-Flash(Nano-Banana))相比,我们的方法可以避免内容漂移,显著改善延迟,并生成视觉上令人愉悦的编辑,同时保持高水平的保真度。项目页面:https://openimaginglab.github.io/InstantRetouch/。
英文摘要:
Language-guided photo retouching aims to adjust color and tone while preserving geometry and texture. Recently, diffusion-based retouching shows a superior visual quality, but often struggles with both fidelity issues due to its generative nature and efficiency because of its iterative sampling process. In this work, we propose an efficient and fidelity-preserving retouching method using bilateral space manipulation, which is both compact and content-decoupled. Specifically, instead of directly editing pixels or image latents, our model predicts a low-resolution bilateral grid of affine transforms, which are sliced using a learned guidance map and then applied to the full-resolution image. This approach yields both high fidelity and improved efficiency. To retain strong priors of a pretrained generative model, we distill a multi-step diffusion model into our bilateral grid framework using Variational Score Distillation, complemented by a prompt alignment loss to guide instruction-following behavior. Additionally, we introduce a new benchmark and evaluate our method across multiple dimensions: fidelity, instruction following, and efficiency. Compared to the latest retouch methods, like Gemini-2.5-Flash (Nano-Banana), our method can avoid content drift, significantly improve latency, and generate visually pleasing edits, while maintaining a high level of fidelity. Project page: https://openimaginglab.github.io/InstantRetouch/.
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。