GP22

GP22数据集由韩国科学技术高级研究院创建，专注于汽车设计领域，特别是汽车侧面轮廓的设计特征。该数据集包含1480张来自37个品牌和10个汽车细分市场的侧面轮廓图像，每张图像均标注了设计特征，如车身、车身侧面、驾驶室、日光开口、车轮和轮胎等。创建过程中，设计师参与了数据标注，确保了设计元素的准确位置。GP22数据集的应用领域主要在于辅助汽车设计师进行设计决策和创意发展，通过量化数据支持设计直觉，减少设计过程中的时间和资源浪费。

The GP22 dataset was developed by the Korea Advanced Institute of Science and Technology (KAIST), with a focus on automotive design, specifically the design features of vehicle side profiles. This dataset comprises 1,480 side profile images from 37 automotive brands across 10 vehicle segments, where each image is annotated with design features including the vehicle body, side body, cabin, daylight opening, wheels and tires, among others. During the dataset's creation, professional designers participated in the annotation work to ensure the accurate positioning of these design elements. The main applications of the GP22 dataset are to assist automotive designers in making design decisions and advancing creative ideas, support design intuition via quantitative data, and reduce time and resource waste throughout the design process.