Rapidata/open-image-preferences-v1-more-results

该数据集旨在创建10K文本到图像的偏好对，用于评估图像生成模型在不同图像类别中的表现。通过使用来自fal/imgsys-results的提示，并根据复杂性和质量进行演化，然后让社区为每个提示的两个生成图像进行偏好标注。Rapidata收集了来自全球的170,000个个人偏好，每个图像对大约有10个偏好标注。标注者被要求根据描述选择他们更喜欢的图像，每个图像对被展示给10个标注者，图像的位置随机打乱。数据集的格式包括原始较小结果集的格式和更详细的Rapidata格式。

This dataset aims to create 10K text-to-image preference pairs to evaluate the performance of image generation models across various image categories. Using prompts from fal/imgsys-results, which are evolved based on complexity and quality, the community was asked to annotate preferences between two generated images for each prompt. Rapidata collected over 170,000 individual preferences from people around the globe, with roughly 10 preference annotations per image pair. Annotators were asked to choose the preferred image based on the description, with each pair shown to 10 annotators and image positions shuffled randomly. The dataset format includes both the original smaller result set format and a more detailed Rapidata format.