BIKED

BIKED数据集是由麻省理工学院机械工程系的研究人员创建的，包含4500个由专业设计师设计的自行车模型。该数据集提供了包括装配图像、组件图像、数值设计参数和类别标签在内的详细设计信息。BIKED旨在支持数据驱动的自行车设计方法，通过机器学习技术探索和理解自行车设计空间，以及通过算法生成新的自行车设计。数据集的应用领域包括自行车设计的自动化、性能优化和市场需求的预测，旨在解决自行车设计领域的创新和效率问题。

The BIKED dataset was developed by researchers from the Department of Mechanical Engineering at the Massachusetts Institute of Technology (MIT), and contains 4,500 bicycle models designed by professional designers. This dataset provides comprehensive design information including assembly images, component images, numerical design parameters and category labels. The BIKED dataset aims to support data-driven bicycle design approaches, explore and understand the bicycle design space through machine learning techniques, and generate novel bicycle designs via algorithms. Its application areas include automation of bicycle design, performance optimization and market demand forecasting, with the goal of addressing the innovation and efficiency issues in the bicycle design field.