关键词: 玻璃与陶瓷, 造型设计, 卷积神经网络, 人工智能

Abstract:

In the field of arts and crafts design, the summarization and design of glass, ceramics, and other objects are the most crucial stages in the entire design process, and improving the efficiency of design is a challenge that all designers need to address. Therefore, by employing the Convolutional Neural Networks (CNN) algorithm in artificial intelligence technology, an intelligent model is used to classify the shapes in an intelligent way. The CNN model is configured with the activation function ReLU, maximum pooling method, Softmax processing for the output results, and the cross-entropy loss function. The Adam optimization method is applied to train the model, resulting in a highly accurate model with minimal loss. With the assistance of this model, the designed outcome can be quickly identified whether it follows the modern design principles and promotes the market application of ceramics and glass, thus facilitating better and faster engagement in the subsequent production phase.

Key words: glass and ceramics, modeling design, convolutional neural networks (CNN), artificial intelligence (AI)