ziyu3141/rf_newtrain_1_42

这是一个包含图像审美评估的数据集，其中包含了文件的名称、审美分数、瑕疵分数、错位分数、总分等。每个图像还对应有瑕疵热图和错位热图，以及错位标记、是否不平整等元数据。数据集还包含了首选和非首选的图像，以及修改前后的图像信息。数据集分为训练集，可用于图像审美和图像质量评估的机器学习模型训练。

This dataset contains image aesthetics assessment, including file names, aesthetics scores, artifact scores, misalignment scores, overall scores, artifact heatmaps, misalignment heatmaps, misalignment token labels, unevenness, preferred and unpreferred images, revised images, unrevised IDs, and preference labels. The dataset is split into a training set, which can be used for training machine learning models for image aesthetics and image quality assessment.