VisImages
收藏资源简介:
VisImages数据集是由浙江大学CAD&CG国家重点实验室创建的,专注于可视化出版物中的图像数据。该数据集包含12267张图像,来源于1397篇IEEE InfoVis和VAST的论文,每张图像都配有详细的标题和边界框注释,总计35096个可视化元素。数据集通过一个综合的可视化分类法进行组织,旨在支持可视化领域的文献分析和自动化任务,如可视化分类和检测。VisImages不仅提供了可视化设计的新颖性和复杂性,还通过其详尽的注释支持了人工智能在可视化研究(AI4VIS)中的应用,为研究者提供了一个宝贵的资源,以探索和理解可视化领域的发展趋势和设计实践。
VisImages Dataset was created by the State Key Laboratory of CAD&CG at Zhejiang University, focusing on image data from visualization publications. This dataset contains 12,267 images sourced from 1,397 papers published in IEEE InfoVis and VAST conferences. Each image is accompanied by detailed captions and bounding box annotations, totaling 35,096 visualization elements. Organized via a comprehensive visualization taxonomy, the dataset is designed to support literature analysis and automated tasks in the visualization field, such as visualization classification and detection. VisImages not only showcases the novelty and complexity of visualization designs but also supports the application of artificial intelligence in visualization research (AI4VIS) through its exhaustive annotations, serving as a valuable resource for researchers to explore and understand the development trends and design practices in the visualization domain.

- 1VisImages: A Fine-Grained Expert-Annotated Visualization Dataset浙江大学CAD&CG国家重点实验室 · 2022年



