A Study on Improving the Accuracy of Classification Models through Property Image Generation

Abstract

In the case of business development by learning information about properties, such as recognizing scratches on the floor or wall, it is difficult to find a certain pattern in the teacher data, which makes it difficult to prepare teacher data of sufficient quality and quantity. Therefore, in this study, we investigate how to improve the accuracy of the classification model by generating pseudo-images from a small number of samples to improve the quality and quantity of the teacher data.