Indoor display handicraft art system by deep learning framework in colleges and universities

The study aims to analyze the application of the interior display handicraft art system based on the deep learning framework in colleges and universities. An indoor space handicraft display positioning recognition model based on Convolutional Neural Network (CNN) is implemented aiming at the error problem in the display of indoor space handicrafts in colleges and universities. Firstly, based on the mobile network, the handicraft data in the indoor space is collected through edge nodes. After that, the data collected by the terminal is analyzed through edge computing. Finally, the data is fed into AlexNet for training. The accuracy, loss value, and response time of the implemented models are further compared and analyzed. The results show that in the analysis of the accuracy, the implemented model is the highest, reaching 82.89%. The convergence rate is the fastest, and finally converges at 0.06 compared with other neural network algorithms. In the response time comparison, when the test distance of the proposed algorithm increases to 15m, the average response time is reduced to 0.19s/m, which is the shortest compared with other algorithms. The application of deep learning to the indoor display handicraft art system offers significant improvements in accuracy and feature extraction effectiveness. This research provides an experimental foundation for the intelligent design of item displays within interior spaces, laying the groundwork for future advancements in this field.