漫画图像编辑中的推理时间轨迹优化
Inference-time Trajectory Optimization for Structure-Preserving Manga Image Editing
AI总结：
本文提出一种在推理时适应预训练图像编辑模型的方法，通过仅使用输入漫画图像来调整模型，以提高对漫画图像的编辑效果，同时减少计算开销。
摘要：
我们提出了一种推理时间适应方法，该方法利用仅输入图像本身来定制预训练的图像编辑模型。尽管近年来预训练图像编辑模型取得了进展，但这些模型在漫画图像上往往表现不佳，因为它们主要在自然图像数据上训练。然而，由于计算成本和版权限制，对大规模模型进行重新训练或微调通常是不现实的。为了解决这个问题，我们的方法在推理时略微校正生成轨迹，使得在空提示下能够更忠实重建输入图像。实验结果表明，我们的方法在仅带来微小计算开销的情况下，始终优于现有基线方法。
英文摘要：
We present a lightweight, training-free trajectory correction method that adapts a pretrained image editing model to each input manga image using only the input itself. Despite recent progress in pretrained image editing, such models often underperform on manga because they are trained predominantly on natural-image data, while re-training or fine-tuning them on manga is costly and raises copyright concerns. Many manga image editing tasks encountered in practice are structure-preserving, requiring local details to be modified while the input's global composition is retained. To support this common editing setting, our method corrects the early editing trajectory by anchoring it to an empty-prompt reconstruction trajectory. Experiments indicate improved performance in the main text-removal setting, while qualitative examples suggest better composition preservation in screentone synthesis. With FLUX.1 Kontext on an RTX A6000, the method incurs 11% runtime overhead and 0.1% peak-memory overhead; an additional runtime measurement with Qwen Image Edit 2509 on an NVIDIA H200 shows only a 0.1% increase.
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。
