面向图像合成任务的感知伪影定位
Perceptual Artifacts Localization for Image Synthesis Tasks
AI总结：
该研究针对图像合成中感知伪影需人工校正的问题，构建了含10168张图的多任务伪影标注数据集，训练出可跨任务、少样本适配新模型的定位模型，还提出放大修复流水线校正伪影并拓展了多项下游应用。
摘要：
深度生成模型的最新进展推动了各类任务中逼真图像的生成，但这些生成图像往往在特定区域存在感知伪影，需要人工校正。本研究对跨多种图像合成工作的感知伪影定位（Perceptual Artifacts Localization，PAL）开展了全面的实证研究。我们构建了一个包含10168张生成图像的新型数据集，每张图像都标注了覆盖10种合成任务的逐像素感知伪影标签。基于该数据集训练的分割模型可有效定位多类任务中的伪影，此外我们还验证了该模型仅用少量训练样本就能适配未见过的模型的能力。我们进一步提出了一种创新的放大修复（zoom-in inpainting）流水线，可无缝校正生成图像中的感知伪影。通过实验分析，我们阐明了多项实用的下游应用，例如自动伪影校正、无参考图像质量评估以及图像异常区域检测。数据集和代码已公开。
英文摘要：
Recent advancements in deep generative models have facilitated the creation of photo-realistic images across various tasks. However, these generated images often exhibit perceptual artifacts in specific regions, necessitating manual correction. In this study, we present a comprehensive empirical examination of Perceptual Artifacts Localization (PAL) spanning diverse image synthesis endeavors. We introduce a novel dataset comprising 10,168 generated images, each annotated with per-pixel perceptual artifact labels across ten synthesis tasks. A segmentation model, trained on our proposed dataset, effectively localizes artifacts across a range of tasks. Additionally, we illustrate its proficiency in adapting to previously unseen models using minimal training samples. We further propose an innovative zoom-in inpainting pipeline that seamlessly rectifies perceptual artifacts in the generated images. Through our experimental analyses, we elucidate several practical downstream applications, such as automated artifact rectification, non-referential image quality evaluation, and abnormal region detection in images. The dataset and code are released.
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。
