In the context of the upgrading of China's manufacturing industry, the design of industrial products is particularly important. Deep learning can be involved in automotive design activities through a variety of functions. We can use deep learning technology to design the appearance of cars through steps such as establishing a database, extracting feature values, drawing three-dimensional curves, unifying model dimensions, and analyzing errors. In this paper, the appropriate deep learning algorithm is selected through experiments to build a car image generation model based on StyleGAN2. The car image is generated by StyleGAN2 to achieve the purpose of designing car styling, and the full label database and the generation model are combined to form a car styling design tool based on deep learning, and the usability of the tool is tested through actual use. The new tool uses StyleGAN2 to generate car images for the purpose of designing car styling. The conclusion is that the function of artificial intelligence assisting designers to design automobile styling can be realized, which provides a new path for the styling research and development of other products.
Read full abstract