过于鲜艳而不真实？生成色彩保真度的基准评测与校准
Too Vivid to Be Real? Benchmarking and Calibrating Generative Color Fidelity
提供了一个可复用的色彩保真度评测基准（CFD）与学习式指标（CFM），并利用 CFM 的注意力引导无训练的空间-时间引导尺度调制，可直接用于诊断和修正 T2I 模型在写实风格下的过饱和、过对比偏差。
原文依据 we present Color Fidelity Dataset (CFD) and Color Fidelity Metric (CFM) for objective evaluation of color fidelity in realistic-style generations
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论近期文本到图像（T2I）生成的进展大幅提升了视觉质量，但生成在视觉上接近真实摄影的图像仍然具有挑战性。部分原因在于现有评测范式的偏差：人工评分与偏好训练得到的指标往往偏好饱和度和对比度被夸张的鲜艳图像，这使得即使提示词要求写实风格图像，生成结果也常常过于鲜艳而不真实。为解决该问题，我们提出了色彩保真度数据集（CFD）与色彩保真度指标（CFM），用于对写实风格生成中的色彩保真度进行客观评测。CFD 包含超过 130 万张真实与合成图像，并带有有序的色彩真实感等级；CFM 采用多模态编码器学习感知色彩保真度。此外，我们提出了一种无需训练的色彩保真度精修方法（CFR），在生成过程中自适应地调制空间-时间引导尺度，从而增强色彩真实性。CFD 支撑 CFM 进行评估，而 CFM 学到的注意力进一步引导 CFR 精修 T2I 保真度，形成一个用于评估与提升写实风格 T2I 生成色彩保真度的渐进式框架。数据集与代码见 https://github.com/ZhengyaoFang/CFM。
英文原文摘要 1275 字符
Recent advances in text-to-image (T2I) generation have greatly improved visual quality, yet producing images that appear visually authentic to real-world photography remains challenging. This is partly due to biases in existing evaluation paradigms: human ratings and preference-trained metrics often favor visually vivid images with exaggerated saturation and contrast, which make generations often too vivid to be real even when prompted for realistic-style images.To address this issue, we present Color Fidelity Dataset (CFD) and Color Fidelity Metric (CFM) for objective evaluation of color fidelity in realistic-style generations. CFD contains over 1.3M real and synthetic images with ordered levels of color realism, while CFM employs a multimodal encoder to learn perceptual color fidelity. In addition, we propose a training-free Color Fidelity Refinement (CFR) that adaptively modulates spatial-temporal guidance scale in generation, thereby enhancing color authenticity.Together, CFD supports CFM for assessment, whose learned attention further guides CFR to refine T2I fidelity, forming a progressive framework for assessing and improving color fidelity in realistic-style T2I generation. The dataset and code are available at https://github.com/ZhengyaoFang/CFM.