An Automatic GUI Generation Method Based on Generative Adversarial Network

As a technique applied with artificial neural networks, deep learning is widely used in the field of image recognition. However, a lack of available datasets leads to imperfect model learning. By analysing the data scale requirements of deep learning and aiming at the application in graphical user interface (GUI) generation, it was found that the collection of GUI datasets is a time- and labour-consuming project. This makes it difficult to meet the dataset needs of current deep learning networks. To solve this problem, we propose the user interface generative adversarial network (UIGAN), a semi-supervised deep learning model, to produce a large number of reliable GUI datasets. By combining a cyclic neural network with a generated countermeasure network, UIGAN can learn the sequence relationship and characteristics of data, make the generated countermeasure network generate reasonable data, and then expand the selected Rico dataset. Relying on the network structure, the characteristics of collected data can be well analysed, and a large number of reasonable data can be generated according to them. After data processing, a reliable dataset for model training can be formed, which alleviates the problem of dataset shortage in deep learning.