Optimization of Advertising Design CAD Model Driven by Deep Learning and Interactive VR Display

Abstract

The aim of this study is to enhance the intelligence of the computer-aided design (CAD) model for advertising through the application of deep learning (DL) technology while refining its interactive visualization in virtual reality (VR). By precisely adjusting the model parameters, the performance of each model on advertising design data is evaluated. The results show that the CNN_LSTM model is significantly superior to traditional CNN and LSTM models in accuracy, false alarm rate, F1 evaluation index, and execution speed, especially in dealing with unbalanced data sets and a few categories. This innovative method not only improves the intelligent classification and recognition ability of advertising design but also provides more efficient technical support for real-time interactive display in a VR environment. Through the optimization of the DL-driven advertising design CAD model, this study has achieved a technical breakthrough in the field of advertising design, laying the foundation for intelligent development in this field in the future.