Application of Deep Learning in Computer Aided Jewelry Product Design

Abstract

Computer-aided jewelry design is one of the commonly used expression methods in the stage of modern jewelry design. Due to the rapid growth of computer-aided manufacturing technology, using computer to design and making jewelry samples through jewelry rapid prototyping equipment has developed into an important link for jewelry enterprises to develop products. In this study, deep learning (DL) and computer-aided design (CAD) technologies are applied to jewelry product design, and a three-dimensional modeling algorithm of jewelry products based on convolutional neural network (CNN) is proposed, so as to realize intelligent computer-aided jewelry product design. The test results show that the algorithm is simple, excellent in performance, capable of extracting the object contour and accurately obtaining the object boundary. The number of test samples began to increase, and the image recognition errors of different recommended methods showed a downward trend. However, compared with the traditional support vector machine (SVM) and back propagation neural network (BPNN), the image recognition error of this method is obviously lower. Parametric design can not only reduce the designer's time cost, but also achieve many design effects that traditional jewelry modeling can't achieve.