TEdBench++

该数据集是一个经过修订的基准，包含了120个条目，专门针对文本图像处理任务。它解决了先前存在的矛盾，并扩展了编辑任务的种类，包括多种具有挑战性的图像操作类型，如多条件编辑、对象/概念移除、风格迁移以及复杂的替换。该数据集的规模为120张图像，其任务是进行文本图像操作与编辑评估。

This dataset is a revised benchmark containing 120 entries, specifically tailored for text-based image manipulation tasks. It resolves prior inconsistencies in existing resources and expands the range of editing tasks, including various challenging image manipulation types such as multi-condition editing, object/concept removal, style transfer, and complex substitution. The dataset consists of 120 images, and its core task is text-based image manipulation and editing evaluation.