Billyyer/RICO-Screen2Words-Rename

Screen2Words是一个提供移动屏幕截图总结（即移动屏幕的图像描述）的数据集，基于RICO图像数据库构建。它包含了截图的多种形式，包括普通截图、图标、语义截图，以及对应的描述和应用的元数据信息，如应用类别、下载次数、评分等。这个数据集可以用于开发多模态自动化工具，自动生成屏幕总结、基于语言的界面检索、增强屏幕阅读器等。

Screen2Words is a dataset providing screen summaries (i.e., image captions for mobile screens) based on the RICO image database. It includes various forms of screenshots, such as regular screenshots, icons, semantic screenshots, as well as corresponding captions and metadata about the apps, such as app category, number of downloads, ratings, etc. This dataset can be used for developing multimodal automations for automatic screen summarization, language-based UI retrieval, enhancing screen readers, and more.