A Survey on Automatic Code Generation Using Image Mock-Ups

Abstract

The idea is to use mock-ups as a way of designing the user interface of a complex program. The most complicated part of the software project is arguably writing the code to model the mock-up. So rather than spending time coming up with an algorithm to take care of the details, we use automatic code generation based on an action, objects relation, an optional default screen view. Deep learning models which transform image mock-ups of graphical user interface (GUI) to a program. It taken place by four steps object detection, object cropping, object recognition, program builder. This are used to build the websites, mobile application and software for the particular use. While in the stage of object identification the cropped components are labeled by CNN trained model.