FILKOM and FCS UB Lecturers Conduct Intelligent System Research, Creating Generative AI for Batik

A research was carried out by the Intelligent Systems Laboratory, Faculty of Computer Science (FILKOM), Universitas Brawijaya (UB). Research led by Dr. Eng. Novanto Yudistira, S.Kom., M.Sc. The research team also involved Dr. Eng. Irawati Nurmala Sari, S.Kom., M.Sc. from FILKOM, and assisted by several members from the Faculty of Cultural Studies (FCS) UB. One of FILKOM students involved in working on this project was Octadion, and it was then continued by Daffa Izzuddin, who made this project as the topic of his thesis.

This research focuses on developing a system based on generative artificial intelligence (Generative AI) which allows users to produce batik designs according to their wishes. This system uses “prompt to batik” technology, where users can provide a textual description of the desired batik pattern, and generative AI will automatically produce a batik design that matches that description. Starting with the collection process in 2022. The generative AI research process will be implemented in 2023.

The research team collected and analyzed a dataset containing various traditional batik motifs from all over the archipelago. A generative AI model was trained using this dataset to recognize and understand the patterns, colors and distinctive elements of various traditional batik patterns. The model training process is carried out in UB Tesla A100 AI Center server, which provides the high computing capacity which need to train generative AI models. Meanwhile, FILKOM UB server is used for the inference process, where users can give prompts and receive batik designs produced by the system.

“Users can provide a specific prompt, such as batik with a blue parang motif with a modern touch, and the generative AI system will produce a design that fits that description. The research results show that this system is capable of producing batik designs that not only maintain the authenticity and beauty of traditional Indonesian motifs, but can also be adapted to the user’s individual preferences and creativity,” said the Team Leader who is familiarly called Yudis.

Yudis explained further that this technology has the potential to enrich and preserve Indonesian batik culture through digital innovation. It can also be used as a useful tool for designers and the batik industry to create new creations more efficiently and purposefully. The use of generative AI in this research shows a high ability to produce unique and on-demand design variations, making the batik making process more dynamic and adaptive to market trends and needs. In addition, this technology opens up opportunities for wider creative exploration in batik art, enabling the creation of new motifs that remain rooted in tradition but are able to answer modern aesthetic demands. [pon/yudis/UB PR/ Trans. Iir]