New carpet pattern design with deep learning

The fact that digitalization touches every aspect of today’s world has often been realized with the ideas industry and projects carried out for development industry. While Artificial Intelligence shows its effective activities in many areas it can also support designers pre-production design. Within scope this study, is aimed to produce new product designs by using image data most demanded products obtained from a carpet manufacturing company Deep Convolutional Generative Adversarial Networks. Before performing model training images manufacturer company, generation performances ready-made set two different Python libraries, Keras PyTorch, were compared. As result, was determined performance PyTorch library generating higher quality, built real repeated. synthetic produced presented manufacturer, bring role designer position guides design models designer’s experience knowledge process, which complex stochastic process before production.